A Road Self-Guided Hardware-Based Demo System

This paper introduces a road self-guide demo system using a full hardware algorithm implemented on FPGA in order to detect two painted lines on the road serving for direction navigation. This system focuses on processing images captured from the complicated environment including noise and different lighting conditions by a single camera in the real-time. The architecture allows to develop a hardware algorithm CORDIC from a software algorithm Hough Line Transform to detect a line shape model. For real-time processing approach, the Dual-ports SRAM combined with the pipeline architecture embedded in its read/write operations are implemented. This system was implemented and successfully proved in an Altera DE2-70 FPGA Development Kit. To prove the real-time processing of hardware system, the timing diagrams are analyzed. In addition, the comparison between the hardware implementation and programming one is conducted to illustrate the benefit of hardware solution on the performance.