形态滤波器在交通视频检测中的应用

Application of Morphological Filters to Traffic Video Detection

介绍交通视频检测的基本过程 ,分析了交通视频图像的特征 ,研究了三种形态滤波器———中值滤波器、普通形态滤波器、顺序滤波器对交通视频图像的处理能力 .顺序滤波器的使用依赖于二值图像中噪声与目标的分布参数P .提出了选择交通视频图像分布参数P和滤波窗大小的基本方法 .实验证明 ,选用合适参数P的顺序滤波器能够大大减少计算量 ,快速地处理交通视频的二值图像 ,基本满足在各种气象条件下的实时采集交通数据的要求 .

The basic process of traffic video detection was i ntroduced and the ch aracteristics of transportation video image were analyzed. By studying the traff ic video image processing ability of three morphological filters, namely, the me dian filter, the ordinary morphological filter and the order filter, it is found that the utilization of the order filter depends on the noise and the distributi on parameter P of two-value images. A basic method of selecting P and th e filter window width of traffic video images was also proposed. Experimental results sh ow that, by selecting suitable P value, the order filter can predigest the c alcu lation and can make the image processing speed faster. The filter can ultimately meet the real-time need of traffic data collection under various weather condit ions.