Image Blur Based on DCT and Clustering Algorithm
Jie Zhang · Jisuanji fangzhen · 2012
Study the problem of increasing the image sharpness.Image enhancement technology has been playing a vital role in image processing.For the traditional image enhancement algorithm easily leads to the probem such as block effects,in order to improve the image quality after image enhancement,the paper proposed a multi-scale image enhancement algorithm.based on discrete cosine transform and mean clustering.First,the basic pieces of the image were merged into a larger matrix,resulting in a wider range of image enhancement factor to avoid the non-image correlation between small.Then according to the different characteristics of different bands,the adaptive enhancement coefficient was used to make the image enhancement effect of different regions more dependent on their own characteristics.Finally,by means clustering of image smoothing,the overall image brightness and contrast were improved to obtain the final image enhancement.Simulation results show that the proposed algorithm effectively improves image clarity,and the image synergistic effect is more smooth.