基于敏感区域多结构元素形态学边缘检测算法

在边缘检测算法的应用中,很难同时兼顾图像处理效果和处理速度,为此,提出了一种新的边缘检测算法,将边缘处理集中在感兴趣的图像区域中.该算法利用梯度算子获得图像中的敏感区域,再构造多种结构元素,结合形态学梯度和OTSU分割法检测敏感区域的边缘.应用于沥青路面裂缝图像检测,实验结果表明,与其它边缘检测算法相比,该算法不仅具有很好的边缘提取能力,而且具有很强的抗噪能力.在保证处理效果的同时,也保证了处理速度,有很高的实用性和推广性.

Edge Detection Algorithm for Sensitive Region Based on Morphology of Multi-Structural Element

In the practical application of edge detection, it is difficult to make a balance between image processing effect and processing speed. In order to solve this problem, a novel edge detection algorithm is presented which the interested regions of the image are processed only. The proposed algorithm employs the gradient operator to gain the sensitive regions in the image. After that, combining the morphology-gradient operator with OTSU segmentation algorithm, it constructs multi-structural element to detect the edges in sensitive regions. The presented algorithm is applied to the asphalt pavement crack detection. The experimental result indicates that compared with other edge detection algorithms, it has not only a good ability to extract the image edge, but also a strong ability to suppress the noise in the image. Moreover, it can provide good image processing effect and processing speed.