International Journal of Computer Applications in Technology (53 papers in press)

Special Issue on: "Advances in Computer Aided Design and Digital Entertainment"
Guest Editors: Prof. Xiaopeng Wei and Prof. Qiang Zhang, Dalian University, China Dr. Xiaopeng Sun, Liaoning Normal University, China Prof. Jun Wang, The University of New South Wales, Australia

• Proximity Queries using Separating Bounding Volumes
  by Yi Wang, Wenhui Li, Yi Li, Zihou Ge 
  Abstract: This paper presents a fast method for determining the collision or proximity information for 3D models using Separating Bounding Volumes (SBVs). SBVs are the tightest volumes in the optimum separating direction of two models, so that they can help achieve high culling efficiency in collision detection process. In order to achieve real-time performance even for complex, non-convex models, we design an algorithm to speed up the computation of SBVs by using the support vector machines (SVM) and a robust scheme to conduct penetration search under the framework of SBVs. As opposed to most techniques, our SBVs based proximity query method requires no specific knowledge of the object’s geometry or topology, or any pre-processing computations. The examples indicate that by our method, the meaningful local penetration information can be computed for complex models and challenging intersection scenarios within a few milliseconds.
  Keywords: collision detection; bounding volumes; support vector machines.
   
• A Stylistic Human Motion Editing System Based on A Subspace Motion Model
  by Gengdai Liu, Zhigeng Pan, Xi Cheng 
  Abstract: In this paper, a stylistic human motion editing system is presented. Compared with many previous works that either edit low-level motions or learn styles from large motion data sets, our system concentrates on how to edit high-level styles of motions and obtain style from only a motion pair. A new low-dimensional motion model is introduced into this system. Based on this motion model, style is defined quantitatively as a low-dimensional subspace of motion data and can be obtained easily from a motion pair. Our system provides several style editing tools and accessorial modules based on this subspace motion model. Using these tools and modules, animators can not only translate the style of the original motions but also transfer and add styles between two motions. Our system is intuitive and convenient for animators.
  Keywords: style; character animation; motion capture; editing system
   
• Q-Cartoon: A Novel Cartoon Creating System
  by Xi-Dao LUAN, Yu-Xiang XIE, Ling-Da WU, Jia-Xu WU, Li-Qiong DENG 
  Abstract: How to generate cartoon in real time is a hot spot in cartoon area. This paper proposes a real time cartoon creating system called “Q-Cartoon”. Q-Cartoon is a text-to-cartoon creating system, which is quite different from traditional key frame generation method. It synthesizes the techniques of Chinese word segmentation, script analyzing and controlling, and databases, and can make text into fancy cartoons. Q-Cartoon has many advantages, such as real-time generation, small storage space, convenience, easy manipulation, etc. and has a wide foreground in the area of education, entertainment, etc. This system explores a creative way for reusing multimedia data, and with the development of related research, its potential values and applications can be widely extended.
  Keywords: Cartoon Generation, Chinese Word Segmentation, Script Analyzing, Script Controlling, Multimedia Data Reuse
   
• Semi-automated Human Body 3D Animator
  by Emmanuel Tsekleves, John Cosmas, Cenkai Yang 
  Abstract: The creation of realistic 3D human body animations is one of the most challenging research topics of investigation in the domain of the 3D animation, as it constitutes an extremely time consuming task that requires significant technical and creative expertise. The motivation of this research project is to develop a tool to simplify and automate the creation of human three dimensional movements by enhancing a 3D graphics creation tool. The proposed tool semi-automates the creation of human character animation, by specifying the starting and ending states and movement styles of a human avatar and the transitionary motion state between them. The tool combines the traditional key-frame technology with the footstep motional circle method and it uses the transition method to create complete human character animations.
  Keywords: computer aided animation; digital entertainment; semi automated animation; computer graphics.
   
• A 3D Shape Classifier with Neural Network Supervision
  by Zhenbao Liu, Jun Mitani, Yukio Fukui, Seiichi Nishihara 
  Abstract: The task of 3D shape classification is to assign a set of unordered shapes into pre-tagged classes with class labels, and find the most suitable class for a newly given shape. In this paper, we present a 3D shape classifier approach based on supervision of the learning of point spatial distributions. In this classifier, we first extract the low-level features by characterizing the point spatial density distributions of 3D shapes, and train one feedforward neural network to learn these features by examples. The Konstanz shape database was chosen as the test database, and we grouped the classified objects into two sets, the training set and the test set, which each had an approximately equal number of shapes. We trained the network with the training set, and evaluated the accuracy rate on the test set. We also compared this classifier to k nearest neighbors classifier for 3D shapes. This approach can be used to classify 3D shapes and enhance the performance of the existent 3D shape retrieval methods.
  Keywords: 3D shape classification; 3D shape classifier; point spatial distributions; neural network supervision.
   
• Real-time Digitized Shadow Play Performance Method Based on Multi-point Interactive Controlling Method
  by Cheng Yao, Jun Xiao, Tong-qiang Guo 
  Abstract: Shadow play is one of the intangible cultural heritages in China and was popular in most areas of our country. But few people watch the play now because of the increasing popularity of new media. To attract more people’s attention, we offer a new digitized performance method based on multi-point interactive controlling. The method maps transforms of a small group of controlling points to character controlling points accordingly. We define mapping rules that decrease the complexity of controlling with fingers. Even the performer with no skills can perform well in our runtime system. With this method the performers can perform digitized shadow play effectively with no limitation of the traditional shadow play resources.
  Keywords: digitized shadow play, digital character, multi-point interactive control, multi-point mapping
   
• Heuristic Nondominated Sorting Genetic Algorithm-II for Satellite-Module Layout Optimization
  by Yi XU, Can Li, Yanjun Shi, Hongfei Teng 
  Abstract: Layout optimization of satellite-module belongs to a kind of complex engineering layout and combinatorial optimization problem. Pareto-based multi-objective evolutionary algorithms, such as NSGA-II, have shown much success for solving the combinatorial optimization problems. Because of the satellite-module layout belonging to complex layout optimization problem, the other strategies also have to be adopted to solve the problem effectively. In this study, we solve the problem by a new approach based on NSGA-II. Firstly, we apply a diversity degree method to evaluate the dominance (or inferiority) of solutions in the same front instead of the crowding-distance method, aiming at finding more layout design schemes. Secondly, we introduce a filter technique of Pareto solution set to preserve the elite individuals more effectively. Thirdly, we adopt three kinds of operators (i.e., multi-parent arithmetic crossover, uniform and Gaussian mutation) to improve search performance. Finally, we give two heuristic layout strategies (position exchange for two components and component jump) in the optimization process. The above algorithm is called Heuristic Nondominated Sorting Genetic Algorithm-II (HNSGA-II). In the numerical experiments of the simplified satellite-module layout design, the experimental results show the feasibility and the effectiveness of the proposed HNSGA-II.
  Keywords: Layout optimization; Satellite-module;Multi-objective evolutionary algorithm; NSGA-II; Pareto solution; heuristic Nondominated sorting.
   
• Calibration and Segmentation free 3D Modeling from images based on GPU
  by Bo Shu, Xianjie Qiu, Zhaoqi Wang 
  Abstract: In this paper, we present an automatic image based modeling method based on shape from silhouettes which does not need any user interactions of camera calibration or image segmentation. Under circular motion constraints, using an iterative optimization of graph cuts and Powell’s direction set minimization method, we can label object’s visual hull and minimize silhouette coherence, which is the difference between projected visual hull and the background subtracted silhouettes. This minimization process can finally converge to both accurate camera parameters and smooth visual hull. Using commodity GPU’s strong parallel computation ability, our approach is not only automatic but also efficient which can produce realistic 3D models directly from several images. The method can be widely used in digital entertainment, digital heritage, digital visual effects and etc.
  Keywords: Shape from Silhouettes; calibration; segmentation; graph cuts; Powell’s method
   
• A Two-stage Multimodal Speaker Location-aware Approach in Pervasive Computing
  by Ruo-gui Xiao, Tong-qiang Guo 
  Abstract: Location-aware computing is one of the key issues in applications such as pervasive computing and intelligent video surveillance. In this paper, we propose a two-stage multimodal approach to locate the active speaker in intelligent environments. In the first stage, human voice is captured as audio cue to find the approximate orientation of current speaker. In the second stage, the color feature of mouth region is extracted as visual cue to detect continuous mouth motion that identifies the active speaker. The speaking recognition is conducted by a well-trained Hidden Markov Model based on color feature of mouth region during continuous motion. Experiments show that the proposed multimodal approach is effective for speaker localization in intelligent indoor environments.
  Keywords: location-aware; multimodal; time delays of arrival; hidden markov model.
   
• Watermark based Digital Fake Image Detection
  by Wei Sun, Wei Lu 
  Abstract: Due to the cheapness and convenience of desktop image editing software, such as Photoshop, GIMP etc., it is becoming popular of altering the contents of digital images, which is defined as image fakery. Whether or not, fake images are full of our daily lives. What we can do is to prove it. Therefore, in this paper, a detection method is proposed to examine that whether an image is fake or real. In order to locate the altered parts, a watermark based pre-processing method is introduced. If the image is fake, the extracted watermark is different from the original one. The proposed method not only can detect whether an image is fake, but also can locate the fake area. Experimental results also show that the performance of the method is encouraging.
  Keywords: Image Fakery; Fake Detection; Digital Watermarking.
   
• Efficient reconstruction from architectural drawings
  by Ting LI, Bo SHU, Xianjie QIU, Zhaoqi WANG 
  Abstract: In this paper, we present an efficient method for reconstructing 3D building-models from electronic architectural drawings (EADs). EADs are matched via recognizing axes and elevation. After recognizing candidate-contours (CCs) of architectural components in one floor plan, the method incorporates the results into recognizing the neighbour floors, which are achieved by a mapping algorithm. CC classification is proposed to reduce the interaction for 3D reconstruction. A parameterized 3D model database is used to improve details of reconstructed results. Our method requires some interaction as guidance. The experimental results show that ordinary users could create a 3D building model which consists of over 100 components in several minutes.
  Keywords: contour mapping; contour classification; parameterized 3D model DB; reconstruction.
   
• An Efficient Simplification and Real-time Rendering Algorithm for Large-scale Terrain
  by Lai Kang 
  Abstract: This paper proposes a Triangulated Irregular Network (TIN) based method for simplifying and real-time rendering of large-scale terrain. The whole terrain is firstly divided into terrain blocks during the terrain partition process, and for each block, feature points are picked out to generate a TIN model. Then the hierarchical multi-resolution models of the terrain can be created in a bottom-to-up mode. Compared with regular models, the efficient TIN model used in this paper reduces the triangle numbers more greatly and remedies cracks between terrain blocks more easily by sharing boundary. When rendering the terrain, whether a terrain block is to be rendered or not depends both on its isotropic error metric and the result of view frustum culling, and an additional synchronous thread is employed to dynamically load or release the terrain data. These rendering strategies could improve the rendering efficiency because they increase the throughput of GPU and avoid load imbalance among CPU, GPU and I/O. Experimental results show that the system performs visually smooth rendering of large-scale terrain scenes with a fine quality at an interactive frame rate.
  Keywords: real-time rendering; 3d terrain; multi-resolution; triangulated irregular network; quad-tree
   
• Content-Based Organization, Analysis and Retrieval of Soccer Video
  by Junqing Yu, Yunfeng He, Kai Sun, Xingming Ouyang 
  Abstract: For bridging the semantic gap, specific-domain approaches are employed in our system, a fully automatic and computationally efficient framework for the indexing, browsing and retrieval of soccer video is proposed. A hierarchical organizing model based-on MPEG-7 is introduced to effectively represent high-level and low-level information in soccer video. The model’s attributes have been instantiated through a series of multi-modal processing approaches. To effectively retrieve highlight clip of soccer video, adaptive abstraction is designed based on the excitement time curve, and an XML-based query scheme for semantic search and retrieval is proposed based on the hierarchical model. For retrieval, XQuery, a XML query language, is employed in the proposed querying framework. The proposed framework employs visual, audio, and cinematic features, and can be scalable to different preferences and requirements. Its efficiency, effectiveness, and the robustness of the proposed framework have been demonstrated over our extensive experiments.
  Keywords: hierarchical organization model; soccer video; event detection; semantic retrieval
   
• A Script Engine for Realistic Human Motion Generation
  by Jun Xiao, Tongqiang Guo 
  Abstract: This paper proposes a framework of script engine for realistic character animation based on motion capture (MoCap) database. Our script engine produces realistic human animation in either offline mode or online mode. In offline mode, users make motion scripts, describing the motion properties such as type, order, and details, by a common text editor or a graphical user interface provided by our system. Then the scripts are decomposed into sequential commands for retrieving relevant motion clips from a MoCap database and generating final animation sequence. In online mode, users employ keyboard, mouse or other peripheral input equipments to activate motion commands and obtain synthesized movement sequence in real-time. Furthermore, users can define their own motion elements table and scripts format suitable for various MoCap data sets. In experiments, we built two applications based on this script engine and verified their capability on two different MoCap data sets respectively. The experimental results show that this script engine framework can produce robust results and be used as a human motion engine in various applications, such as computer games, movies, sports simulation and virtual reality.
  Keywords: script engine; human motion; motion capture data.
   
• A Reactive and Protective Character Motion Generation Algorithm
  by Xi Cheng, Gengdai Liu, Zhigeng Pan 
  Abstract: By combining motion capture and dynamic simulation, character animation obtains satisfactory and realistic visual effect, and the character can also respond to unexpected contact forces from the environment. In order to generate physics based character animation in real time, this paper introduces an artificial neural network which is trained on the samples offline and predicts a subset of the return-to motion capture (mocap) database online to narrow the search. In the simulation phase, we propose a simple but effective balance detection method for the characters in locomotion. In addition, our human biomechanics based controller can drive the characters taking several balance recovery actions when they lose balance, and taking the protective actions to reduce the injury on their bodies when they fall onto the ground. Compared with the previous character motion generation methods, the analytical and visual result of the character animation indicates that our algorithm can run in real-time and can be used in reactive character games or other interactive applications.
  Keywords: character motion; dynamic simulation; balance detection; artificial neural network; protective behaviours
   
• Extraction and analysis of the speech emotion features based on multi-fractal spectrum
  by Qirong Mao, Yongzhao Zhan 
  Abstract: In order to reflect the change of pitch, speech energy and frequency of emotional speech signal more accurately, in this paper, a calculating method for the multi-fractal spectrum of speech envelop, frequency wave and pitch contour of speech signal is presented. According to these spectrums, 18 novel speech emotion features are proposed. In order to test the validity of the multi-fractal emotion features for the speech emotion classification, the clustering performance of these new features are analyzed. In addition, by using multi-fractal features proposed in this paper, traditional acoustics emotion features and the features combined the multi-fractal features with the acoustics features respectively, recognition experiments were done for 7 emotions (happiness, sadness, surprise, anger, fear, disgust and neutral). The experiment results demonstrate that, compared with the features of the other two kinds, the features combined the multi-fractal features with the acoustics features show the highest classification performance. And by using the multi-fractal emotion features only, the recognition accuracy attains to 76.7%. So the speech emotion features proposed in this paper are effective for the speech emotion recognition, and they can be used as the complementary features to the traditional acoustics emotion features for the speech emotion recognition.
  Keywords: speech emotion recognition; multi-fractal theory; emotion feature; clustering analysis
   
• A multi-scale triangular mapping and its applications in variation scale scrambling
  by Li-Ping Shao, Zheng Qin, Hong-Jiang Gao, Bo Liu, Huan Li 
  Abstract: To solve the variation scale image scrambling problem, a multi-scale triangular mapping and its inverse mapping are presented in this paper. To increase the scrambling space variation, an mton mapping is proposed. Combined with the proposed mton mapping, scrambling method based on multi-sclae triangular mapping has a flexible variation scale scrambling ability. To compare with conventional image scrambling methods, it can scramble any image with numerous different scales in pixel position space to change the pixel position relevance or in pixel colour space to change the pixel grey value relevance or in both pixel position and colour space to change both the pixel position and grey value relevance. The experiments show the proposed multi-scale based scrambling method flexible ability in scrambling image, low cost in scrambling and recovering image, wonderful scrambling performances by comparing with conventional scrambling methods.
  Keywords: triangular mapping; multi-scale; variation scale; image scrambling; image security; rectangle image.
   
• Combining Location and Feature Information for Multimedia Retrieval
  by Hong Pan, Yi Yang 
  Abstract: Since nowadays most of the entertainment resources are in the form of multimedia, better management of multimedia data will be very helpful for it. In this paper, we propose a cross-media retrieval method of heterogeneous multimedia data by which the query examples and the returned results can be of different modalities, e.g. to query audios by an example of image. Because it is very common that the heterogeneous multimedia data of the same semantics always exist jointly in many domain and application specific databases and the completed semantics is expressed by all the co-existing multimedia objects which can be of different modalities, there is abundant location information which can be used to better understand the multimedia semantics. Taking location and feature information into consideration, an affinity analysis approach is proposed to analyse and fuse the information carried by the co-existing multimedia objects so as to learn the semantic correlations among the heterogeneous multimedia data and perform cross-media retrieval. We also propose an active learning method of Relevance Feedback from the users to make the search model more accurate.
  Keywords: Heterogeneous Multimedia Data, Relevance Feedback, Active Learning, Cross-media Retrieval.
   
• Local resampling for patch-based texture synthesis in vector fields
  by renjie chen, Ligang Liu 
  Abstract: In order to synthesize distortionless texture on surfaces, we develop a direct and accurate approach for local resampling in vector fields, and then use the approach to synthesize textures on 2-D manifold surfaces directly from a texture exemplar. Regular-grid patches produced by the local resampling are used as building blocks for texture synthesis. Then texture optimization and patch-based sampling are generalized to synthesize texture directly in vector fields. The first scheme can create texture of higher quality; however the second scheme is faster and simpler and works well for a wide range of textures. Users can control the vector field on the mesh to generate textures with local variations including the orientation and scale. Many experimental results are presented to demonstrate the ease of use and reliable results provided by our system.
  Keywords: Texture synthesis, vector field, local resampling, texture optimization, patch-based sampling
   
• Facial Expression Recognition Based on Combined HMM
  by wen chuanjun 
  Abstract: In this paper, we put forward a new recognition method which combines HMM and Mixture HMM by KNN, while Mixture HMMs are the models that traditional HMMs are trained by any two kinds of training samples. The whole recognition procedure is to make final class judgment by KNN while HMM basic decision is different from Mixture-HMM confirmatory sentence. At the same time, some improvements for KNN decision criterion and HMM interactive rule are made to enhance classification ability. Experiments show that the performance of this combined method exceeds that of HMM based only.
  Keywords: facial expression recognition; combined HMM; Mixture-HMM; HMM interactive rule; KNN.
   
• Synthesis of Cartoon Fire based on Hand-drawn Samples
  by Yi Gu, Jinhui Yu, Ren Peng 
  Abstract: In this paper we present an approach for synthesizing and editing cartoon fire series based on hand-drawn samples. Our approach first extracts outlines of hand-drawn cartoon fire series and, based on the central reference skeleton, synthesizes fire of arbitrary length by high level control on flame skeletons. With our approach users may edit cartoon fire series to achieve effects such as fire spreading and fire blown by wind from side.
  Keywords: Cartoon fire; image synthesis; computer animation
   
• A Novel Automatic Motion Capture Data Recognition Method Based on Statistics Learning and Subspace
  by Jian Xiang, HongLi Zhu 
  Abstract: In this paper, we propose a motion recognition method based on motion capture data. In order to recognise motion type, a generalized Isomap nonlinear dimension reduction based on radius basis function(RBF) is used to project original motion data into low-dimensional subspace. Then some motion type classifiers are learned for each human's joint in subspace. Then we use ensemble learning to enhance learning results. Experimental results show that our methods are effective for 3D human motion recognition .
  Keywords: motion capture; RBF; recognition; subspace, Isomap.
   
• Dynamic clothing collision resolution using PSO
  by Wenhui Li, Yi Wang, Yi Li, Qi Jiang 
  Abstract: This paper presents a novel stochastic method for dynamic clothing simulation and collision detection by using Particle Swarm Optimization (PSO) algorithm. Firstly, we convert the collision detection problem to a multi-objective dynamic environment optimization problem. Then use a density-based clustering PSO to find optimal or near-optimal solutions. During the searching process, we also put forward a restrictive fly algorithm to prevent particles from flying out of the solution space. Experimental results demonstrate that PSO has good performances in our optimization problem. Besides, with Active Pair Searching and BVHs, our method can achieve high detection ratio and balanced performances at interactive rates.
  Keywords: cloth simulation; collision detection; particle swarm optimization
   
• An Interactive Motion Editing Method for Human Animation
  by Mofei Song, Yaoye Zhang, Zhengxing Sun 
  Abstract: This paper presents an interactive motion editing method for human animation. Our method combines sketch modelling with the animation technique based on motion capture. Through sketch, users are able to rapidly create a simple 3D skeleton model without any constraints to express exaggeration. In order to make full use of captured motion data, motion retargeting is used to adapt the original motion from the standard model to the model obtained from sketch and the motion is adjusted to meet different constraints according to the motion model automatically. For supporting expression of intention, our method also provides users direct manipulation of the key parts of the model and then uses inverse kinematics to update the gesture. Finally, a new motion is created by smoothly propagating the motion displacement to neighbour frames in terms of linear interpolation. Experimental results prove the effectiveness of the proposed method.
  Keywords: motion edit; skeleton modelling; motion retargeting; inverse kinematics;trajectory editing
   
• A New Approach for Texture Classification in CBIR
  by Shengju Sang, Mingxia Liu, Jing Liu, Qi An 
  Abstract: In the field of Content-Based Image Retrieval, the semantic understanding of textures has long been a difficult problem, especially the texture classification. The present approaches for texture classification are mostly based on mathematical parameters, which can not break through the semantic obstacle between the visual features and human understanding of textures. This paper proposed a new approach for texture classification, which adopted ten words describing textures in natural language. Texture features of an image are extracted by DWT(discrete wavelet transform), and then classified through both BPNN(back propagating neural network) and SVM(Support Vector Machine). Experimental results show that the approach of texture classification for natural texture is feasible.
  Keywords: CBIR; texture classification;DWT; BPNN; SVM; WPT.
   
• An Emotion Generation Model Based On Random Graphs
  by Chaogang Wan 
  Abstract: Emotions are an indispensable aspect in harmonious human-computer interaction and artificial intelligence. In this paper, we present an algorithm for generating emotions for a 3D virtual human and its emotions are shown with 3D facial expression animation. According to the theory of emotions, we use the PAD (Pleasure-Arousal-Dominance) emotion space to define affective elements which are used for the emotion generation. These elements include exterior stimulus which are affective speeches and facial expressions from people in the real world, and the previous emotions and moods of the virtual human. We use a stochastic graphical model, which is a stochastic recurrent network, to describe the relationship among the emotion, the mood and the personality. Finally, we build a mapping from the emotion space to the facial expression space with a competitive learning network. The facial expression is vividly shown by a 3D virtual human head. Using our experimental results, we analyze the influences of mood and emotion decay on emotion generation. The results have shown our emotion generation model works effectively and meets the basic principle of human emotion generation.
  Keywords: emotion generation; PAD emotion space; stochastic graphical models; 3D facial expression animation.
   
• Improve Matching Results for Structure from Motion Problems
  by Chunmei Duan, Xiangxu Meng, Chenglei Yang 
  Abstract: In this paper we propose a method to improve feature point matching results for structure from motion problems, which includes techniques for promoting the correspondence quality using epipolar geometry, fast decision from relative positions of matches and local geometry constraints. We start with feature points extracted from each image and initial set of matches obtained by cross correlation and relaxation principle, then several strategies exploited to detect outliers. Minimizing back projection error in projective space is the traditional method to discard outliers. However, the retrieved projective structure is needed first. Our method is performed before projective reconstruction phase and only two views of scene are needed. Experiments have been carried out and the results demonstrate the efficiency of proposed strategies.
  Keywords: feature point matching; fundamental matrix; epipolar geometry; 3D reconstruction
   
• Multi-chessboards localization based on FCM and Radon Transform algorithm
  by Zhou Dongsheng 
  Abstract: Due to the simple and striking structure, Chessboard is widely used as the camera calibration pattern. Locating the chessboard in an image is a prerequisite step during the calibration procedure, especially under the situation that there exists more than one chessboard with random relative positions in an image. A robust recognition method based on fuzzy C-means clustering (FCM) and Radon Transform (RT) algorithm is presented to solve the problem above. Experiments show that the method is effective to locate and distinguish the chessboards with a large and complicated background from each other. The proposed method also provides a possibility to calibrate a camera with one shot.
  Keywords: camera calibration; chessboard localization; FCM; Radon Transform
   
• Contour recovery of tablet calligraphy characters
  by Xianjun Zhang, Jun Dong 
  Abstract: A lot of Chinese calligraphy characters were carved on the stones. Most of the tablet calligraphy characters captured from these stones were damaged by people or eroded by natural environment. It is useful to recover the damaged contours of the tablet calligraphy characters automatically. A scheme of recovering the contour of tablet calligraphy characters is proposed, which includes two phases, pre-processing and contour recovery. In the pre-processing phase, filtering and segmentation are implemented to remove the noise and locate the tablet calligraphy characters respectively. Binarization and morphological algorithm are used to obtain binary image without small holes within the character. In the contour recovery phase, the first step is stroke extraction, which is to decompose the character into strokes and to obtain the layout of strokes. The eight-field searching method is applied to extract the contour of the strokes. The next step is to recover the contour of calligraphy strokes based on Fourier descriptors with less Fourier coefficients. The recovered strokes are combined together to form the calligraphy character based on the layout as the last step. The experimental results are encouraging, which shows the recovered and constructed calligraphy character is smooth and aesthetic.
  Keywords: calligraphy; contour recovery; contour extraction; contour smoothing; tablet.
   
• ASaaSI: an approved Architecture for SaaS Service Composition
  by Yuchang Jiao, Shijun Liu, Xiangxu Meng, Chenglei Yang 
  Abstract: With the development of Internet and web2.0 technology, SaaS (Software as a Service) has generally been accepted as main way of the software service delivery. However, the SaaS applications are isolated, which could not follow the rapid change of the business requirements. In this paper, after analysis of the business flow and IT resource available based on two general cases, SBM and APO, an approach to integrate SaaS services together was proposed. We give the SaaS service integration architecture and solve several critical problems in the presentation layers, function layer and data layer. Based on the architecture for integration we proposed, developers can easily combine the SaaS offerings together to meet new business scenario. At last we implemented the supporting system and put it into real practice in a large-scale car manufacturing factory in china.
  Keywords: Service Composition; SOA; SaaS; EWidget Model; SURROGATE; Service Coordination; Flow Combination.
   
• Distribution System Insulator Monitoring Using Video Surveillance and Support Vector Machines
  by Sreerama Murthy Velaga, Dusmanta Mohanta, Sumit Gupta 
  Abstract: The electric power distribution system can be conceived as arteries of the power system, carrying power to the consumers. With the phenomenal growth of such networks into remote areas even in developing countries such as India, the development of Substation Monitoring Systems (SMS) with augmented features for substation automation purposes has become one of the potential areas for research. The ongoing SMS automation schemes mostly incorporate protection of the distribution system from different faults based on monitoring parameters such as voltage and current. But damaged insulators of the distribution system also affect the system performance very strongly in terms of reduction in voltage and flow of leakage currents. The proposed technique is intended for condition monitoring of insulators on an overhead distribution system using Video Surveillance (VS) in contrast to the conventional practice of on-site physical detection. It employs Support Vector Machines (SVM) to incorporate expert evaluation for obtaining coherent conclusions regarding condition monitoring of insulators based on analysis of the images obtained. The most significant contribution described in this article is proposed scheme that can be used as an effective tool for real-time substation automation.
  Keywords: Support Vector Machines (SVM); Substation Monitoring System (SMS); Distribution Automation System (DAS);Video Surveillance
   

• B-spline Surface Fitting to Random Points with Bounded Boundary Conditions
  by Kuo-Jen Chen, Yao-Chen Tsai, Jiing-Yih Lai, Wen-Der Ueng 
  Abstract: Two algorithms for the surface fitting of random points have been developed. The first algorithm is applied when the fitted surface passes through four given boundary curves, while the second algorithm is applied when an extended surface bound is to be achieved. The fitted surface is essentially a B-spline surface where the control points and the parametric values corresponding to each of the measured points are unknown. A generalized optimization algorithm, including the initial parameterization and parameter optimization, is proposed for these two cases. A plane and a Coons surface are created for the above two cases, respectively, to determine the initial parametric values. An objective function including an error function and an energy function is then developed for the optimization of the unknown parameters. Specific governing equations are developed which consider the effects of the boundary constraints for each of the cases. A thoughtful discussion of the optimization algorithm is provided and the feasibility of the proposed method is demonstrated.
  Keywords: surface fitting; random points; B-spline surfaces; surface fairing; boundary constraints.
   
• 3-D Transient FE Analysis and Movement Simulation for the modelling of Magnetic Levitation
  by M'HEMED RACHEK, Mouloud Féliachi 
  Abstract: This paper presents a three-dimensional transient numerical modelling of an electromagnetic levitation problem with taking into account the movement. The governing eddy current equations given in terms of magnetic vector and electric scalar potentials are solved with the help of 3-D Finite Element Method (FEM) tools implemented under Matlab environment. The coupled eddy current-mechanical model is carried out through the computation of electromagnetic force acting on the moving part, and solving the motion equation to get the displacement. In order to obtain an accurate solution of the magnetic field, the reverse Cuthill-McKee ordering is used to reduce the bandwidth of the matrix and then ensure fast convergence of the Incomplete Cholesky Bi-Conjugate Gradient (ICBiCG) solver. The calculation of the electromagnetic force is performed by volume integration of the Lorentz force. The body movement at each step displacement is simulated using the proposed Fast Partial Remeshing Technique (FPRT) in which only the moving part is remeshed without any changes on the mesh topology and on the number of degree of freedom. An aluminium cylindrical plate located above two cylindrical coils is presented as an example.
  Keywords: Transient eddy current, Magnetic force, Finite element method, 3-D movement simulation.
   
• Characterizing the aggregate structure variations in asphalt concrete during compaction using computer automated imaging techniques
  by Kasthurirangan Gopalakrishnan, Naga Shashidhar 
  Abstract: The aggregate structure or internal structure of Asphalt Concrete (AC) mixes, which refers to the content and the spatial and directional distribution of air-voids, aggregates and asphalt, evolves during the process of compaction, which is an important step in the construction of AC pavements. During the process of compaction, aggregates coated with asphalt are brought together by pressure and kneading action while the volume of air trapped within it is reduced. This study focused on applying computer automated two-dimensional imaging techniques and near-neighbour distance methods to quantify the variations in AC internal structure during the process of compaction. This was done by studying the inter-particle distances in cross-sectional images of AC specimens compacted using a linear kneading compactor at different air voids. The results of the study indicate that the changes in inter-particle distances during compaction could be measured using two-dimensional image analysis of a sawed section of a sample. Quantifying AC internal structure directly through techniques developed in this study and establishing its relationship to pavement performance will be very beneficial in designing pavements to have the appropriate internal structure to maximise the performance. This will provide the foundations to building more durable and long-lasting pavements.
  Keywords: Asphalt Concrete (AC); image analysis; near-neighbor distance methods; Delaunay triangulation; Voronoi diagram; compaction
   
• Point Cloud Simplification with Bounded Geometric Deviations
  by Hao Song, Hsi-Yung Feng 
  Abstract: This paper presents a new method for point cloud simplification, which is to reduce the number of data points included in a dense point cloud sampled from the surface of a 3D object (often via a high-speed laser scanning/digitizing system). The proposed method searches for a subset of the original data such that the maximum geometric deviation between the original and simplified data sets is below a specified error bound. The underlying principle of simplification is to partition the original data set into piecewise point clusters and represent each cluster by a single point. The two main tasks of the associated search algorithm are thus to iteratively update the partition of clusters and to efficiently evaluate the resulting geometric deviations. It starts with an initial simplified data set which is likely to exceed the error bound. The cluster representation is then iteratively updated by increasing the number of clusters, or equivalently the points included in the simplified data set, until the error bound constraint is met. The algorithm then proceeds to a point-removing stage to delete redundant point clusters and reduce the size of the simplified data set as much as possible. Effectiveness of the proposed method has been demonstrated through several case studies using practical point cloud data sets.
  Keywords: Point cloud data; Simplification; Data reduction; Geometric deviation; Clustering
   
• Computer-Aided Generation of Fixture Configuration Design Using Polychromatic Sets
  by Ping Ji, Francis K. H. Lau, Lili Jiang, Min Li, Zhongbin Li 
  Abstract: This paper presents a new approach to the generation of fixture configuration design in manufacturing. First of all, the concept of polychromatic sets is introduced. Then the polychromatic set approach to the generation of fixture configuration design is proposed. This approach first identifies the relationship between fixture elements, then derives a contour matrix and its corresponding Boolean matrix. Furthermore, feasible paths and final alternatives to the fixture configuration design are found using the theory of polychromatic sets. A fixture configuration design example is used to illustrate the approach.
  Keywords: polychromatic sets, fixture configuration design, computer aided fixture design
   
• Model reference approach to a force feedback joystick in PC video games
  by Jia-Sheng Hu, Mi-Ching Tsai 
  Abstract: This paper aims to develop a viable impedance control based on a model reference scheme for power-aided joysticks in recreational video game applications. The impedance control approach can endow the joystick with interactive force feedback to enrich the amusement in PC video games, through which the game designer can easily specify the desired torque responses of the controlled joystick. The focus of this research is to establish haptic sensing of a virtual object with an observer to estimate the load torque information without using any torque sensors. The controller design and constraints are discussed thoroughly, and experimental results are given to evaluate the performance of the proposed control scheme.
  Keywords: impedance control; load torque observer; force feedback control; model reference.
   
• A Computer-Aided Design Tool for the Analysis of Electrical-Connector Transmission Lines
  by Wei-chen Lee, Sedig Agili 
  Abstract: This paper presents the calculation of the differential impedance, the velocity of propagation, and the worst-case near-end crosstalk coefficient of the 2D cross sections of transmission lines by using a computer-aided design tool based on the finite difference method. The 2D cross sections of the transmission lines embedded in this tool are commonly seen in electrical connector design, but not in electrical circuit design of printed circuit boards or semiconductors due to their different fabrication processes. Several transmission line designs are studied and the important results are presented. Some charts that illustrate the relationship between the electrical parameters and the geometries of the transmission lines are shown, which allow the designers to perform a quick initial design. Further calculation can be easily performed by using the tool.
  Keywords: impedance; electrical connectors; transmission lines.
   
• Modelling PLC Characteristics for Resource Allocation
  by Lucien Ngalamou, Leary Myers 
  Abstract: This paper presents an approach that models PLC (Programmable Logic Controller) characteristics for automatic dimensioning and allocation in process control. The automation of an industrial process is invariably faced with the decision of the optimal resources for implementation. It is important to find a way of selecting the most efficient programmable logic controller that is suitable to facilitate a particular control process. Variables such as the number of inputs or outputs, program memory size, data memory size, auxiliary memory, timers and counters have to be given the requisite consideration. The method undertaken considers the use of ladder diagrams and XML in the process of modeling PLC characteristics which are organized in form of a database. Criteria for PLC selection are used to provide a list of PLCs or embedded controllers that best fits the control process envisioned. This approach can be used as a decision making tool by machine design companies, small manufacturing companies and consultant engineers facing with the correct choice of PLCs to be deployed in a particular control project.
  Keywords: PLC characteristics; programmable logic controllers; ladder diagrams; XML models; databases; resource allocation; process control.
   
• Integrated manufacturing system of high pressure FRP pipes
  by Jiazhong Xu 
  Abstract: In order to realize industrialized manufacturing of epoxy FRP pipes with high efficiency, the manufacturing system which can accomplish winding, internal heating curing and extraction processes by only one machine tool was developed. The winding motion and logic control are undertaken by an embedded motion controller, and the electronic-gearing is used to achieve synchronous motion control of the spindle and the carriage. A thermo-chemical model during cure was established, which is used to predict the degree of curing and the amount of the heat generated. An analysis based on the finite element technique was developed to simulate the curing process of FRP pipes. Based on the numerical results, the optimized cure cycles with the heating and cooling steps were developed. The curing control system composed of embedded industrial computer and PLC was designed to undertake curing, cooling and blowing control, and the extraction process and its control system are introduced. The performance of the FRP pipes manufactured by this system was tested, and it is proved that the applying of internal heating curing process can improve the production efficiency and the quality of high pressure FRP pipes.
  Keywords: high pressure FRP pipes; winding; internal heating curing; integrated manufacturing.
   
• Wavelet Series Based Iterative Learning Controller Design For Industrial Robot Manipulators
  by Selvaraj Gopinath, Indra Kar, R. K. P. Bhatt 
  Abstract: In this paper, a wavelet series based learning controller has been proposed for the tracking control of robot manipulators. Wavelet series approximation is used to approximate the desired and actual trajectories of the system into finite number of wavelet coefficients. A learning controller is designed in wavelet domain which forces the wavelet coefficients of the actual output approach to the corresponding wavelet coefficients of the desired trajectory, which are known constants, such that the tracking of robot manipulator is achieved. Numerical experimentation results are presented, and the paper describes the ability of the learning controller designed in wavelet domain over the uncertainties present in the robot manipulator system.
  Keywords: wavelet series; integral wavelet transforms; learning control; robot control; robot manipulators.
   
• Remote Access of Electronic Resources for Thermal Plant Using Mobile Devices
  by Fock-Lai Tan, Sai-Cheong Fok 
  Abstract: Field inspection is an indispensable activity in thermal system management. The real-time exchanges of information between the field engineer and the central management system are vital for accurate and timely fault diagnosis and performance analysis of the thermal processes. This paper proposes a framework that extends a typical intelligent thermal system management to allow field engineers to retrieve vital resources from the centralized management station using mobile devices with limited memory, such as PDAs and cell phones. User input is enabled using a MIDlet running on the mobile device. The inputs are then transferred to a servlet application (developed using the Java 2 Enterprise Edition servlet API on a Tomcat server) to access the electronic resources. The computed results are then transmitted back to the mobile device for display. A case study involving the retrieving of thermodynamic properties of steam and R134a refrigerant through mobile devices is used to demonstrate the functionality of the proposed framework. The computed properties investigated in the case study were retrieved in real-time and within 4% error in the sub-cooled, saturated and superheated regions.
  Keywords: fault diagnosis; thermal management systems; mobile devices; thermodynamic properties; J2ME; remote access; thermal plants.
   
• Quality Optimisation in Engineering Simulations Software
  by George Besseris 
  Abstract: Taguchi methods and fractional factorial designs may aid in assessing the robust development of a software product. This work showcases a simulations problem incurred in the homogenisation process of industrial production of aluminium products. The demonstration is focused on optimising two code quality characteristics for solving numerically the time-dependant, three-dimensional heat loss for a homogenised aluminium slab in queue for hot milling. The equation of heat flow is discretised by a standard parabolic finite-difference method. The response of the two relevant quality characteristics, the total execution time and the total points computed per total execution time is evaluated according to Taguchi’s signal-to-noise ratio. The timestep and the mesh size of the three discretised spatial coordinates play the role of the examined group of performance controlling factors. The analysis is modeled according to a two-level, four-factor L16(215) orthogonal array design retaining all second and third order interactions. The effect, of the order the resulting nested loops are computed, is quantified and subsequently optimised.
  Keywords: software quality; optimisation; design of experiments; engineering simulations; Taguchi methods.
   
• Acquisition Steps of a Remotely Accessible Rapid Prototyping Laboratory
  by Ismail Fidan 
  Abstract: Remotely-accessible laboratories are becoming increasingly popular across a wide range of disciplines. The objective of this project is to develop and disseminate a remotely accessible Rapid Prototyping (RP) Laboratory at Tennessee Technological University (TTU). This laboratory facility is physically located on the TTU Campus, but Internet-based controls, interactive tutorials, and scheduling allows this equipment to be used across TTU, Tennessee Board of Regent (TBR) technology centers and community colleges across the state of Tennessee and eventually nationwide. Since a RP laboratory facility has already been established through a previous National Science Foundation (NSF) grant, this new endeavour increases its accessibility via ubiquitous Web-based communication tools. As such, this enables many more students and instructors to develop their skills set and knowledge base with cutting-edge RP technology. This project has many wide-ranging impacts. Foremost, the work helps promote the awareness of RP technology and improve related teaching and learning paradigms. In particular, adding a remote access capability allows many more users to benefit from state-of-the-art RP technology, thereby better justifying the cost of purchasing and maintaining the overall facility. Furthermore, the successful development of a remote RP lab provides new insights into the strengths and weaknesses of the remote access environments, and therefore provides vital insights for both the design/manufacturing technology and distance education communities alike.
  Keywords: rapid prototyping; remote access laboratories; internet; 3D printing.
   
• A structured process model for conceptual design of mechanisms
  by Chi-Lan Cai, Ren-Bin Xiao 
  Abstract: This paper presents a new structured process model for conceptual design of mechanisms, which is based on axiomatic design (AD), set theory and artificial immune system (AIS). AD framework is used as modelling basis, which includes design domains, decomposition, and zigzagging mapping to achieve hierarchical architecture. The behaviour has been introduced into this zigzagging mapping process to bridge the gap between functional domain and physical domain, which is beneficial to realise intelligent conceptual design. Set theory is used to describe the problem space and solution space to formalise the design, which can be regarded as the underlying representation of design knowledge. Based on this extended mapping process and set theory-based representation, the automated function decomposition technique is presented. An immune optimisation algorithm is introduced into this model to avoid combinatorial explosion during searching for feasible combinatorial solutions. A design case is presented to illustrate the application of this model.
  Keywords: conceptual design; axiomatic design; set theory; immune optimisation algorithm.
   
• QoS Driven Scheduling for Grid: From Application Perspective, Resource Perspective and System Perspective
  by Li ChunLin 
  Abstract: The paper considers QoS-driven resource scheduling from application perspective, resource perspective and system perspective. QoS-driven resource scheduling deals with allocation and scheduling of resources to applications based on the QoS requirements. The objectives of the individual applications, the individual resources, and the system are often in conflict. Application QoS parameters are expressed as a function of the application’s goals. Resource-level QoS parameters depend on the control parameters. Applications running in the system need to be able to specify their QoS requirements and preferably be able to operate over a range of QoS values. In order to resolve the conflicts of the objectives of application and resource providers, the system needs to be aware of application QoS parameters and resource-level QoS parameters to find an optimal relation between user satisfaction and resource utilization. Utility functions are used to express grid user’ QoS requirement, resource provider’s benefit function and system’s objectives. Experimental results show that compared with traditional method, QoS-driven resource allocation and scheduling performance is significantly improved by considering three perspectives.
  Keywords: grid; QoS; application perspective; resource perspective; system perspective; resource scheduling.
   
• Enhanced Bicycle model and Complex Multi Body Dynamics model – A Comparative Study
  by Balaramakrishna Nizampatnam, Krishna kumar R 
  Abstract: Vehicle handling evaluation by using reduced order models is very useful during the pre-design stage. The main aim of this paper is to enhance the well known bicycle model (two degree of freedom model) with different suspension and steering system compliances like roll steer, roll camber, lateral force compliance steer, and steering system compliance and to compare with higher order multi body dynamics models. Different kinematics and compliance coefficients of steering and suspension systems are included into the two degree of freedom model. The frequency domain comparisons of the three models i.e., two degree of freedom model, enhanced two degree of freedom model and multi body dynamics models is carried out with well known four parameter evaluation method of lateral transient response. Time domain comparison has been done by including different compliances to the side slip angles of the wheels. A close match between the sets obtained from the enhanced two degree of freedom model and ADAMS/car model has been observed within the linear range of operation. The effectiveness of the enhanced two degree of freedom model is demonstrated by studying the effect of different suspension and steering system compliances like roll steer, roll camber, lateral force compliance steer, and caster angle variation on vehicle handling potential. Among them, the lateral force compliance steer showed the dominant effect. The interaction analysis showed that there is no significant interaction between different compliances.
  Keywords: Enhanced two degree of freedom model, Kinematics and compliance coefficients multi body dynamic models, two degree of freedom model, Vehicle handling.
   
• Symmetric Reconstruction Algorithms for Incomplete 3D Models
  by Lilit Yenokyan, Armen Zakarian, Pravansu Mohanty, Vahram Avagyan 
  Abstract: Reconstruction of incomplete 3D models of real-life objects and use of these objects in a variety of applications, such as medical, engineering, computer vision, and manufacturing, are of increased interest. In this paper, a new approach and algorithms for 3D model reconstruction, based on the model symmetry, are presented. Every phase of the proposed approach slices the 3D model into layers along one of the axes, then the layers are analyzed individually, and the results of the analysis are combined to produce an overall assessment of the model’s positioning and geometric properties. After the incomplete 3D model is introduced to the system, symmetry type and model parameters are established, and an algorithm is developed for model alignment. Symmetry analysis algorithms that use steps from the slicing procedure are proposed for collecting information on the asymmetric/missing sections of the aligned model on a layer-by-layer basis. The outputs of these algorithms are arrays of points that complete symmetric properties of the layers, which are input to the mesh generation algorithm that generates a triangular mesh based on those points. The methodology developed in this paper is illustrated with the 3D model reconstruction examples. A computation study on the algorithms performance is also presented.
  Keywords: symmetric reconstruction; CAD/CAM; computer aided design; computer aided manufacturing; symmetry analysis; 3D geometry; model alignment; 3D mesh generation; 3D model slicing; planar symmetry; axis symmetry.
   
• Real-Time Hybrid Testing Techniques for Structural Control Devices
  by XiaoQi Chen 
  Abstract: Effective real-time testing of structural control devices relies on a hybrid test system that couples virtual structures under dynamic loading with physical sub-structures or devices in a dynamic test rig. The use of sensors and actuators in a closed-loop feedback system maintains the dynamic equilibrium of the overall system comprising the physical test article and virtual modelled structure. The virtual-real hybrid testing method thus alleviates much of the time and cost associated with full-scale testing and enables tests that would be infeasible without full-scale complete structural tests. This research presents simple, cost-effective and robust hybrid test system, and outlines solutions to the major issues faced in developing any hybrid system. The final system readily accommodates non-linear-multi-degree-of-freedom models and an operating bandwidth of 1 kHz.
  Keywords: hybrid testing, pseudo-dynamic testing, real-time control, virtual structure, simulation, structural control devices
   
• ALLOCATION OF OPTIMAL TOLERANCES IN A MECHANICAL ASSEMBLY USING DE AND NSGA-II
  by BALAMURUGAN CHINNASAMY, SIVAKUMAR K 
  Abstract: This paper discusses a general methodology based on evolutionary algorithms (Elitist Non-dominated Sorting Genetic Algorithm (NSGA-II) and Differential Evolution (DE)) for analysis and allocation of optimal tolerance for an over-running clutch assembly. Minimisation of the manufacturing cost, quality loss function and maximisation of Number of Standard Deviations (ZNSD) are considered together to build a multi-criterion objective function (COF). New hybrid models are developed for obtaining the manufacturing costs of hub, rollers and cage of over-running clutch. The aim of this paper is to find optimal tolerances values which give minimum combined objective function value, minimum manufacturing cost, minimum quality loss function and maximum ZNSD by satisfying stack up conditions using non- traditional optimisation techniques. The results obtained from various techniques are compared and analysed. The proposed techniques give better results than the results reported in literature.
  Keywords: Tolerance allocation; sensitivity analysis; optimisation; mechanical assemblies.
   
• Energy Aware Grid Resource Scheduling: Model and Algorithm
  by Li ChunLin 
  Abstract: Energy efficiency for high performance computing and communication system has recently become an important concern, but most current grid environment doesn’t implement energy aware resource management. The paper proposes an energy aware grid resource scheduling scheme. Energy aware grid resource scheduling optimization is formulated as utility optimization. The goal of the paper is not only to reduce energy consumption, but also to improve the system utility in grid environment, ensuring the battery lifetime and the deadlines of the grid applications. In order to reduce the computational complexity, we decompose the energy aware grid resource scheduling optimization problem into two sub-problems, the interaction between two sub-problems is controlled through the use of the pricing variable. The paper proposes an energy aware grid resource scheduling optimization algorithm. The performance evaluation of our energy aware grid resource scheduling optimization algorithm is conducted by comparing with other related algorithm.
  Keywords: grid resource scheduling, energy efficiency, computational complexity
   
• Region Based Image Registration for Mosaicking
  by Sujoy Roy 
  Abstract: In this paper, a method for image registration using 2D regions as correspondence features is proposed. The proposed method consists of three stages. First, a region segmentation technique is used to find reasonably good regions in image pairs. Second, an adjacency relationship based region matching technique is proposed to find corresponding regions between image pairs. Lastly, a method for finding the affine transformation between regions based on minimum enclosing ellipses is presented. Extensive experiments are performed to demonstrate the efficacy of the proposed method in registering real image pairs subject to large perspective distortions.
  Keywords: image registration; ellipses; mosaics; region segmentation
   
• TWO-DIMENSIONAL LOCALIZATION AND CLEARANCE EVALUATION OF SCREW ROTORS
  by Yi-Chao Huang, Jiing-Yih Lai, Feng-Yung Lin, Yi-Li Guo 
  Abstract: Twin screw compressors are for decades widely used in industry due to advantages over other types of compressors. Due to machining imperfections, deviation from the theoretical profile often causes interference between the rotors. To compensate such deviation, it is necessary to apply clearance between the main and the gate rotors which will still provide minimum clearance with the lowest possible leakage. The aim of this study is to develop a method to analyze clearance distribution between the rotors with the theoretical profiles, modified design profiles and machined profiles. In order to evaluate the position of the contact point on the rotors, a localization algorithm was developed. The clearances were then calculated along the points on the sealing line which was originally derived from the theoretical profile. The validity of the method was finally verified by computer simulation and experiments.
  Keywords: screw compressor, screw rotors, clearance analysis, localization