Real-time edge detection using TMS320C6711 DSP

Abstract

Edge detection is a cornerstone in any computer, robotic or machine vision system. The edge image contains critical information about the scene. In this paper we describe a real-time implementation of edge detection in grayscale image signals using the Canny algorithm. The algorithm is implemented in real time using TI TMS320C6711 imaging developer kit composed of the 6711 DSP and daughter board to communicate with the main DSP. The results show that the execution time is low while the edge results are accurate and thus presenting a suitable algorithm for on-line vision systems applications such as industrial assembly lines, surveillance systems (intruder detection), and on-line hand written document reading. Details of the implementation, required hardware composition, implementation issues, and performance results of real-time real images are presented.