Research On The Lane Detection With DSP And Low-cost CCD

The topic was raised under the background of the industrialization application of the low-cost Automobile Assistant Driving System. Through establishing the lane detection algorithm and structuring the lane detection hardware system, the low-cost lane detection system was established, and in the effectiveness and real-time premise of the lane detection algorithm, it provides a feasible way for the industrial development of the Automobile Assistant Driving System.In this paper, the video processing system was established through utilizing the low-cost CCD camera and high performance DSP development board, and the adaptive threshold segmentation algorithm about road image on the Otsu method was applied in the vision-based lane detection algorithm on the ROI. Then the lane detection algorithm was transplanted on the 32-bit TMS320DM642 DSP model board of the video processing system, and the low-cost lane detection system was established on the video processing system.At last through experiment validated the effectiveness and real-time of the mentioned lanedetection algorithm based on the video.The whole article include five chapters:Chapter oneIn chapter one, we introduce the development status of the visual processing technology,the lane detection algorithm and system based on the video, and a brief description of the process of establishing the algorithm and system. Then we describe the research content and significance, pointing out that the established process of low-cost lane detection system based on video.Chapter twoIn chapter two, we introduce the established process of low-cost lane detection algorithm based on video, include the color space conversion,the gray processing,the adaptive threshold segmentation,the feature point search on the lane about road image under the image plane coordinate system, and the setting of the detection area,a series of horizontal searching zone and ROI on the road image. Also include the corresponding conversion both image plane and body coordinate system about images, the lane curve fitting algorithm based on LMedSquare algorithm, the algorithm initialization and the internal and external calibration theory about the camera etc.Chapter threeIn this chapter, we introduce the established process of low-cost and hardware composition of the lane detection system based on the video processing, include the whole cost calculation and hardware index of hardware system, such as the low-cost CCD camera,32-bit TMS320DM642 DSP development board,LCD screen etc. Also include the initialization of these hardware by CCS integrated development environment, and describe the work process of algorithmin the hardware system.Chapter fourIn this chapter, we introduce the experiment method about the internal and external calibration of the camera, and analysis these experiment results. And we display and explain the single-step processing result for the lane detection algorithm, such as the gray processing effect,the adaptive threshold segmentation processing effect and the lane detection results.Chapter fiveChapter five is the summarization of the article, depicting the main works, researching result and the future research directon.Through the research of the article, the development of the low-cost lane detection system is established, and it provides a feasible way for the industrial development of the Automobile Assistant Driving System, also is the main starting point of the paper. And it lays the foundation for the constantly improving Automobile Assistant Driving System.