摘要
为实现摄像头寻迹小车能在不同赛道中自动、稳定地运行,设计了一种以面阵CCD摄像头作为图像传感器,以MC9S12XS128作为主处理器,利用LM1881将视频信号进行分离处理,并通过TLC5510高速A/D将模拟视频信号进行采集和量化的图像采集与处理系统。在图像预处理中,采用改进的中值滤波算法对图像进行滤波,利用直方图求出动态阀值将图像二值化。此外,在黑线提取中根据不同边线的特点应用了不同的插值算法。实验结果证明,该算法在赛道图像处理中具有良好的鲁棒性、可靠性和实时性,且小车能快速稳定的运行。
In order to make the Camera tracing car can run automatically and stably in different track, we de- sign an image acquisition and processing system using an area array CCD camera as image sensor to collect the image data, using MC9S12XSI28 as the main processor to process digital image, using LM1881 to separate video signals and using the high speed AD TLC5510 gathering and quantify analog video signal. In the image preprocessing, we use the improved median filtering algorithm for image filtering and use histogram calculated the dynamic threshold of two binary images, In addition, we use different interpolation algorithms in the black line extraction according to the different characteristics. Experiments show that, the algorithm has good robustness, reliability and real-time, and the car can run stably and fatly.
出处
《电子科技》
2014年第12期52-57,共6页
Electronic Science and Technology
基金
2013年广东省大学生创新创业训练计划基金资助项目(五邑大学第66号)
关键词
图像处理
图像采集
CCD摄像头
image processing
image acquisition
the CCD camera
