A Histogram Equalization Algorithm Based on Building a Grey Level Binary Tree Dynamically

Abstract Histogram equalization algorithms may produce artifacts when the image is enhanced. Complex filtering algorithms are used for good results, or end-to-end deep learning networks are implemented for image enhancement. The competitive algorithm proposed in this paper uses a binary tree structure to remap grayscale and suppress artifacts. The algorithm can get different degrees of image enhancement results by changing unique variables. The proposed competitive algorithm can also be used to expand the dataset of deep learning tasks. Code and figures are available at https://github.com/F-Quasimo/DEBTHE .