彩色图像的边缘检测

彩色图像的边缘检测算法有很多，但检测结果仍存在细节丢失、抗噪性差的问题。本文提出一种新的彩色图像的边缘检测算法。首先，在RGB空间下，采用矢量梯度算法，分别在三个通道中求出像素的梯度值。然后，采用自适应双阈值算法，对梯度值进行判断，筛选出图像的边缘点。最后，只采用筛选后的边缘点的信息，构造新的模糊矩阵的隶属函数，从而形成新的模糊算法。利用该算法，检测出最终的图像边缘信息。实验证明，该算法得到的图像检测结果，不仅对比度高、边缘连续、细节保存完整，而且具有较强的抗噪性。

The edge detection of color image

There are many edge detection algorithms for color images, but the results of image still have the problems that missing details and poor noise immunity. This paper proposed a new color image edge detection algorithm. Firstly, in the RGB space, vector gradient algorithm was used to calculate the gradient of the pixels in three channels. Then, used an new adaptive double threshold algorithm to judge the gradient values and filter the edge points of the image. Finally, usd the information of the filtered edge points to construct a new membership function of the fuzzy matrix, and used this new fuzzy algorithm to detect the final edge information of the image . Experimental results show that, the image detection results is obtained by this algorithm, it not only has high contrastion, continuous edges, complete details preservation, but also has strong noise immunity.