基于RGB降维模型和小波的彩色图像边缘检测

针对彩色图像边缘检测计算量大的弱点,对彩色图像空间(RGB)进行三基色降维处理,保持图像的各个基色的灰度矩不变,以最大限度地保留原图像的主要特征.用三次B-样条函数作为光滑函数来定义小波函数,通过小波变换对降维后的图像进行边缘检测.实验结果表明,该方法既可以突出细节边缘,也能够保持图像的主要信息,而且比其他彩色图像边缘检测方法平均减少了40%的计算时间.

Edge detection of color images based on wavelet and reduced dimensionality model of RGB

It takes much time to detect edges of color images directly. To resolve this problem, reduce the dimensionality of color image space (RGB), while keep the gray moments of tristimulus values invariable to reserve as much as possible the main characteristics of the original images. Using three step B-spline as the smoothing function to define the wavelet function, the edges of images with reduced RGB dimensionality were detected by the wavelet transform. The experimental results show that this method can emphasize the details of edges, and keep the main information of images,averaging 40% less computing time than other edges detection methods for color images.