通过引入Tsallis熵,提出了一种新的彩色图像边缘检测方法。该方法首先将彩色图像从RGB空间转换到YCbCr空间,分别求出三个子空间中像素在3×3邻域内的最大边缘强度值;然后采用Tsallis熵求取每一个子空间的最佳边缘强度闽值对亮度子空间Y和颜色信息子空间Cb、Cr分别进行边缘提取,最后将三个子空间中获取的边缘点进行融合,从而完成整幅图像的边缘检测。实验结果表明,与采用Shannon熵的边缘检测方法相比,该方法具有更好的灵活性和可行性。
This paper proposed a novel color image edge detection method based on Tsallis entropy to automatically detect edges in color image. First,it rtransformed the image space form RGB to YCbCr, and found the maximal edge intensity in a 3 × 3 neighborhood. Then it used Tasllis entropy to find the optimal intensity threshold for every subspace and extracted the edges in the three subspace. At last,it fused the edges in those three subspace together. The experiment indicates that the proposed method is more feasible and valid than unusual Shannon entropy-based methods.