International Journal of Vehicle Autonomous Systems
Forthcoming articles have been peer-reviewed and accepted for publication but are pending final changes, are not yet published and may not appear here in their final order of publication until they are assigned to issues. Therefore, the content conforms to our standards but the presentation (e.g. typesetting and proof-reading) is not necessarily up to the Inderscience standard. Additionally, titles, authors, abstracts and keywords may change before publication. Articles will not be published until the final proofs are validated by their authors.
Forthcoming articles must be purchased for the purposes of research, teaching and private study only. These articles can be cited using the expression "in press". For example: Smith, J. (in press). Article Title. Journal Title.
Articles marked with this shopping trolley icon are available for purchase - click on the icon to send an email request to purchase.
Online First articles are published online here, before they appear in a journal issue. Online First articles are fully citeable, complete with a DOI. They can be cited, read, and downloaded. Online First articles are published as Open Access (OA) articles to make the latest research available as early as possible.
Articles marked with this Open Access icon are Online First articles. They are freely available and openly accessible to all without any restriction except the ones stated in their respective CC licenses.
International Journal of Vehicle Autonomous Systems (4 papers in press)
Traffic sign recognition using deep learning by Vraj Patel, Joy Mehta, Saurab Iyer, Ankit Sharma Abstract: Recognition of traffic signs is an integral step towards achieving Advanced Driver Assistance Systems (ADAS) as distracted driving is one of the primary causes of road accidents and fatalities. This paper attempts to exploit the capabilities of Convolutional Neural Networks (CNN) to recognise traffic signs under various computational and environmental constraints. The German Traffic Sign Recognition Benchmark (GTSRB) dataset is used for the classification of images. The dataset is subjected to various image processing techniques, such as gray scaling, denoising, filtering, and thresholding, to obtain a generalised model for the recognition of traffic signs. The neural network used here comprises three convolution layers, each followed by a max pooling layer, which further are followed by four fully connected dense layers. The models are trained for 100 epochs with a validation split of 20%. The model performs best with Adam optimiser with a learning rate of 0.001. Keywords: traffic sign recognition; advanced driver assistance systems; German traffic sign recognition benchmark; deep learning; computer vision; image processing; convolutional neural network.
The development of a GPS-based autonomous quadcopter for precision landing on a moving platform by Muhammad Farris Khyasudeen, Norlida Buniyamin, Saaidal Razalli Azzuhri, Muhammad Badri Mhd Noor, Mohamad Hakim Abu Bakar, Mohamad Faizal Abdul Rahman, Nik Izwan Kamel, Aliff Shariffuddin, Iraj Sadegh Amiri Abstract: Unmanned Aerial Vehicle (UAV), especially quadcopter is widely used in search and rescue, mapping, surveillance and infrastructure inspection. The ability of the quadcopter to fly outdoor and return to its take-off point, or another landing point is important as it simplifies the quadcopter operation and reduce the possibilities of the quadcopter from damages and loss. This work proposed a quadcopter that can perform landing operation on a moving landing platform. This system is based on coordinates of both quadcopter and the moving landing platform using Global Positioning System (GPS) and feedback control using radio frequency communication. GPS is used as it reduced the complexity and cost of using another system such as image processing and computer vision which used higher capability processor on the quadcopter, and needed high quality vision sensor. The quadcopter used both the coordinates of quadcopter and moving platform to identify the distance and yaw angle between these two points and land on the moving platform without pilot assistance during the landing sequences. The results show that the accuracy of landing on a shifted platform task based on the performance of the GPS used was 6 to 46%. For best performance, the quadcopter should initially fly at a height of 4 metres. This paper contributes to the advancement of quadcopter precision landing by means of developing method, which use low cost equipment without compensating the accuracy of the process. Another contribution is this paper shows the ability and accuracy of quadcopter to execute precision landing on a moving platform without using image processing techniques and high capability computer on the quadcopter. Keywords: quadcopter; GPS; moving platform; autonomous precision landing; UAV; Radio frequency feedback control; Return-To-Platform.
A new fractional-order sliding mode controller for the cruise control system of automatic vehicles by Sajjad Sheykhi, Hossein GholizadeNarm Abstract: In this paper, a fractional-order sliding mode controller (FOSMC) is proposed for the speed control of automatic vehicles and solve the control problem when tracking a road vehicle. A sliding mode cruise controller is a smart cruise control system used in modern autonomous vehicles. The proposed method is used to control the follower vehicle to follow a reference trajectory transmitted by a leading vehicle. Road safety and driving comfort are among the required constraints when designing such controllers. The proposed controller was validated by modelling via the MATLAB program and then compared to integer-order sliding mode controllers. The simulation results and comparisons demonstrate the effectiveness of the proposed methodologies. Keywords: sliding mode control; fractional-order sliding mode control; cruise control system.
Decoupled 3-D object detector by Moemen Arafa, Ahmed Osama, Mohamed Abdelaziz, Maged Ghoneima, Fernando Garcia, Shady Maged Abstract: This paper proposes an efficient cascaded 3-D object detection architecture. Such an architecture decouples the 3-D object detection pipeline to maximise the use of the inherent advantages of RGB images and LiDAR point clouds in order to perform 3-D object detection while maintaining low computational complexity. Our proposed architecture relies on a cascade of two networks, the first leverages the texture density in images and the maturity of state-of-the-art 2-D object detectors to classify and obtain initial region proposals for objects in the scene. These proposals are fed to a light-weight secondary network that leverages the compactness of bird-eye view point cloud representations to perform orientation and size estimation. The 3-D bounding box proposal is constructed by fusing predictions inferred from both networks, as predictions lie on orthogonal planes. Evaluated on the KITTI benchmark dataset, we show that the proposed method obtains results on-par with more complex end-to-end 3-D detection methods while greatly reducing computational and memory requirements. This work also presents results from the deployment within a perception pipeline, and analyses challenges faced in deployment within a frontal perception pipeline.
Keywords: 2-D object detectors; 3-D object detection; frontal perception; depth representation.