基于水平特征的车辆检测与跟踪

利用加权平均滤波算法对图像进行了滤波处理,去除部分噪声,利用Sobel算子的水平算子对图像的水平边缘进行增强处理,利用最大类间方差法对边缘增强后的图像进行图像分割;提出了利用车辆的水平特征中的列直方图和行直方图对车辆进行检测,根据检测结果得出该方法检测效果良好,即使是复杂环境下的车辆也能得到有效的识别,且对货车的识别效...

Vehicle Detection and Tracking Based on Horizontal Features

The image was filtered by using the weighted average filter algorithm to eliminate the noise.Sobel operator s level operator was used to enhance the horizontal edge of image,and the Otsu method was used to segmentation the image.The vehicle histogram was used to detect the vehicle with horizontal histogram and row histogram.According to the detection results,this method has good detection effect,even in complex environment of the vehicle can also be effectively identified,and the recognition of the truck is also effective.Finally,the method of restricted interesting of region was used to track the vehicle,and the effectiveness of the recognition method and tracking method was verified.