Research On Image Coding Of Lane Departure Warning System Based On DM642

With the constant development of transportation industry, the ownership of vehicles has increased year by year. As a result, a large number of traffic accidents arise. Inspect of the severe situation, the concept of Safety driving assistant system has been first brought up with. What’s more, Lane departure warning system is an important research content of that. As is known, the system is a warning system by sending alert to remind the pilot that vehicle is drifting from the original lane, avoiding the accidents caused by Lane departure.Based on TMS320DM642 chip, a kind of core produced by TI, together with QXD-DM642 DVS development board as hardware development platform, the research of image coding process was conducted under software integrated environment CCS to process the image collected by the LDWS. The program of the development board was debugged on PC in real time under the integrated development environment CCS until the program had been optimized. Then according to the characteristics of on-board FLASH, the program was solidified in the development board, realizing offline operation of the system.This paper systematically studied the principle of the image compression and coding, and the process of image coding. On the basis of system framework, study was focused on the research of the configuration of the encoding chip and analysis of the video stream by I2 C to complete coding circuit design. The image of lane was collected using CCD camera. Then, the collected image was decoded on the image processing subsystem of the QXD-DM642 DVS development board, and the processed image was preprocessed. Next, according to the principle of image coding, the redundant information in the image was removed and the essence of the image was extracted. Finally, the encoded image was displayed on the screen in real time. The real vehicle experiment shows that the system overall technical scheme is feasible, which can effectively collect, process and display the images of video, and the processed image can basically implement the expected purpose.