基于DSP的直线特征提取算法

Research of the linear feature extraction algorithm based on DSP

为解决车道线识别实时性差问题,提出了基于DSP的直线特征提取算法.基于DM6437视频采集系统,通过CCD摄像头采集图像数据,对感兴趣区域提取灰度,以减少实时算法的计算量;在此基础上,通过对有效区域进行改进算法的中值滤波,有效选择噪声和信号点的灰度值进行中值处理,实现道路有效区域边缘保护和消除噪声的功能;再通过提取直线特征边缘检测,根据需要进行Sobel边缘检测,结合直方图筛选、直线连通性分析能有效的平滑噪声并细化边缘,能够达到道路图像的分割、配准和识别的效果,提高算法的准确性和实时性;最后,结合霍夫变换算法,在边缘检测的基础上得到直线参数方程,并且通过先验知识改进Hough变换算法提高算法实时性,准确找出直线位置.

In order to improve the accuracy of the lane line real-time recognition algorithm, this paper presents an algorithm of linear feature extraction based on DSP. Firstly, the algorithm of gray is extracted on the interested region by the CCD camera to reduce the amount of calculation algorithm of real-time images, based on DM6437 video acquisition system. Secondly, the noise and edge preservation are removed by the median filtering algorithm.Then, the function of image segmentation, image registration and target recognition is implemented by the Sobel edge detection algorithm. Finally, in combination with Hough transform algorithm, the straight line parameter equation is obtained by testing the binary image of the straight line.