基于递归细化和局部方差的目标靶边缘检测

Target Edge Detection Based on Recursive Thinning and Local Variance

为了快速精确地提取目标靶的边缘信息,提出了基于递归细化和局部方差的目标靶边缘检测方法。首先采用基于积分图像和Otsu方法的局部方差的边缘快速检测算法对原始图像进行边缘检测,然后建立强边缘宽度-频数图并获取直线宽度,设定直线最小宽度作为递归终止条件,若不满足条件,对原始图像进行遍历八方向法的直线细化处理,然后再重复上述步骤直至满足递归终止条件,得到边缘检测结果。实验结果表明:相比其他常用边缘检测算法,该算法边缘检测准确性和实时性较好,具有较好的抑制噪声能力。

In order to extract the target edges information,an edge detection method based on recursive thinning and local variance is proposed.Firstly the fast edge detection algorithm based on integral image and Otsu method is used to edge detection.And then the histogram of strong edges width-frequency is generated and the line width is acquire,and the minimum width of lines is set as a recursive termination condition.If the detected width is not satisfied that condition,the original image is processed by the line thinning based on the 8-directional image.Then above steps are repeated until the condition is satisfied,getting the result of edge detection.Experiment results show that compared with other algorithms,the proposed algorithm is of great advantage in accuracy and real time attribute with better ability to suppress noise.