基于Meanshift算法的图像边缘检测
Mean Shift-based Edge Detection in Image
本文提出一种以像素的meanshift偏移向量为重要特征,并考虑像素点相互空间关系的边缘检测算法。首先通过meanshift算法计算出图像中各个像素的meanshift最终偏移向量,然后根据meanshift理论以及该偏移量性质,同时引入meanshift向量的方向和长度进行边缘检测。实验表明该算法能精确定位目标轮廓,同时具有良好的抗图像噪声能力。
n edge detection algorithm for image is proposed by referring pixel space relation and features of meanshift vector. Firstly, the meanshift vector for each pixel is calculated; secondly, regarding characteristics of meanshift theory and corresponding vector, the edge detection algorithm is implemented with the information of both length and orientation of meanshift vector. Experiment indicates that the proposed algorithm locates edge accurately and resists nose in image robustly.
赵敏、孙棣华、何恒攀
计算技术、计算机技术电子技术应用
meanshift边缘检测canny
MeanshiftEdge Detectionanny
赵敏,孙棣华,何恒攀.基于Meanshift算法的图像边缘检测[EB/OL].(2009-02-12)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/200902-569.点此复制
评论