Title: Development and optimisation of image segmentation algorithm on an embedded DSP-platform

Authors: Boucif Beddad; Kaddour Hachemi; Sundarapandian Vaidyanathan

Addresses: LTC Laboratory, Faculty of Technology, Dr. Tahar Moulay University, Saida, Algeria ' LTC Laboratory, Faculty of Technology, Dr. Tahar Moulay University, Saida, Algeria ' Research and Development Centre, Vel Tech University, Avadi, Chennai 600 062, Tamil Nadu, India

Abstract: In this work, a very efficient segmentation algorithm based on the analysis of vertical and horizontal histogram by using edge detection method is developed in order to detect the licence plate from a colour vehicle image. VM3224K2 Daughter Kit is used to capture the input image which is to be converted into YUV format and fed to C6713 development kit. The licence plate localisation algorithm is implemented by using MATLAB software and verified for its functionality's. First the in-built functions of MATLAB were replaced by user-defined functions in Simulink Model then translated into Code Composer Studio (CCS) project written in C using an integrated development environment provided by Texas Instrument; also it's compiled and debugged by using CCS, downloaded onto TMS320C6713 DSP-Platform. Finally, VM3224 Daughter Kit is embedded with DSK6713 Kit to display the output image. The proposed algorithm has achieved an accuracy of around 100% and works for both front and rear digital images.

Keywords: image processing; VM3224K2 daughter kit; segmentation; CCS; code composer studio; TMS320C6713 DSK; edge detection.

DOI: 10.1504/IJCAT.2018.10016912

International Journal of Computer Applications in Technology, 2018 Vol.58 No.3, pp.250 - 258

Received: 10 Aug 2017
Accepted: 11 Sep 2017

Published online: 18 Oct 2018 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article