基于中值滤波技术的视频图像边缘检测算法

传统算法抗干扰性能较差，利用人工经验对高低阈值进行设置完成边缘提取，不但会出现伪边缘，同时会产生很多噪声。为此，提出一种新的基于中值滤波技术的视频图像边缘检测算法。依据集合代数原理，通过集合论对几何结构进行描述，完成对视频图像形状与结构的预处理。针对视频图像中某一点的值，对其用该点邻域中不同点值的中值进行替代，对视频图像进行平滑滤波处理。在此基础上，通过设定阈值对某像素点是否是边缘点进行判断，利用最大类间差分技术获取最优阈值，自适应选择合理阈值，以降低假边缘现象的出现概率，提高边缘连续效果。对视频边缘轮廓区域的增强处理，得到有效的边缘检测结果。实验结果表明，所提算法检测结果伪边缘较少，图像边缘检测质量高，噪声少。

Edge Detection algorithm of Video Image based on median filter

The traditional algorithm has poor anti-jamming performance. Using artificial experience to set high and low thresholds to complete the edge extraction, not only pseudo-edges will appear, but also a lot of noise will be generated. To solve this problem, a new video edge detection algorithm based on median filtering is proposed. According to the principle of set algebra, the geometric structure is described by set theory, and the shape and structure of video image are preprocessed. For the value of a point in a video image, the median values of different points in the neighborhood of the point are replaced by the median values, and the video image is smoothed and filtered. On this basis, the threshold is set to judge whether a pixel is an edge point or not, the optimal threshold is obtained by using the maximum difference between classes technology, and the reasonable threshold is adaptively selected to reduce the probability of false edge phenomenon and improve the effect of edge continuity. The enhancement of edge contour region is effective for edge detection. Experimental results show that the proposed algorithm has fewer false edges, higher quality of image edge detection and less noise.