摘要
车道线检测是智能交通中的研究重点之一,论文以DSP为载体,设计并实现了嵌入式车道线检测系统,同时针对于该系统提出了一种新的快速有效的检测算法,该算法结合了特征提取与模型匹配的各自优点,主要通过对图像的采集、自适应、中线提取、车道线匹配和最小二乘法拟合来最终实现。嵌入式系统采用模拟PAL制摄像头采集图像,图像经解码后输入到DSP中得到处理。实验结果表明,该系统车道线检测结果清晰,干扰小,效果良好,并能很好的满足实时要求。
Lane line detection is one of the focuses of the intelligent traffic research, a DSP embedded system was designed and implemented,and aiming at this system, a new fast and effective detection algorithm was proposed. The algorithm which combined the respective advantages of feature extraction and model matching was ultimately achieved through following steps： image capture, adaptation, centerline extraction, lane line matching and least squares fit; and the embedded system captured the image through a simulation PAL video camera which is processed via DSP chip after decoded. Experimental results show that the lane line can be well detected by the system which also can meet the real- time requirement.
出处
《电子测量技术》
2014年第4期83-86,共4页
Electronic Measurement Technology
基金
中国科学院战略性新兴产业科技行动计划(YDJDBHF-2012-001)
江苏省产学研联合创新基金(BY2011173)项目
