Research On Image Enhancement Algorithm Based On Multi-histogram Equalization

With the rapid development of technology,digital image enhancement technology has been widely used.Histogram equalization is one of the main methods of image contrast enhancement.It is used to improve image contrast and achieve the purpose of improving image visual quality and machine recognition of key image features.Histogram equalization is divided into global histogram equalization algorithm(GHE)and local histogram equalization algorithm(LHE).The global histogram equalization algorithm obtains higher contrast by automatically stretching the dynamic range of the image.The GHE algorithm is easy to cause excessive enhancement of the high-frequency histogram and the loss of detail caused by the combination of low-frequency gray levels,and the brightness of the output image cannot be effectively maintained.This paper analyzes the typical algorithms of dual histogram equalization and multi-histogram equalization.Among them,for the double histogram equalization algorithm,it mainly analyzes the BBHE algorithm based on brightness mean and the DSIHE algorithm based on equal area;for the multi-histogram equalization algorithm,it mainly analyzes the RMSHE algorithm based on brightness mean recursion and dynamic histogram equalization DHE algorithm.In order to associate histogram segmentation with image clustering,a dual histogram equalization algorithm based on K-Means image segmentation is proposed.First,K-Means is used to cluster the images to obtain two sub-images.Then,perform histogram equalization processing on the two sub-images respectively,and finally merge the two sub-images.Experimental results show that the algorithm not only has the effect of maintaining the brightness of the image,but also has a better contrast enhancement effect than traditional methods.