关键词: 分数阶微分, 图像增强, 拉普拉斯模板, 均值, 标准差

Abstract: Generally, fractional differential mask doesn’t have flexibility, because it is not filtered with the changes of images’ local information, once it is decided. To remedy the limitations of fractional differential mask, this paper puts forward one way of the enhancement of fractional differential image based on the local feature. The relation between Laplace mask and fractional differential mask based on the 3×3 symmetric fractional differential masks is found out. The fractional differential mask is obtained using the weighted Laplace mask. Based on the relation between local mean and standard deviation, a new method of image enhancement is obtained by improving weighted Laplace mask. This method can execute filter according to the local image feature. It is called as the local feature based image enhancement of fractional differential. Compared with other image enhancement algorithms, the experiments show that the visual results of the way of the enhancement of fractional differential image based on the local feature are the best.

Key words: fractional differential, image enhancement, Laplace mask, mean value, standard deviation