A modular FPGA-based system for vision-based road traffic applications

A Real-time implementation of low-level image processing operations such as filtering and edge detection is not easily achievable unless specific architecture is used. Median filtering and morphological operations are used in vision-based road traffic applications. In this paper a host-based low-cost architecture for these operations is presented. A single processing element of the proposed architecture is only used when the relevant operations are encountered, and the host computer performs the rest of the operation. The result show that the system can provide results in real-time under real-world traffic applications.