Entropy-based Adaptive Histogram Equalization Algorithm

Histogram equalization is one of the most popular techniques in image enhancement. However, due to the shifting of the mean brightness, the quality of the output image is deteriorated with great amount of information loss. This paper proposes a novel modification of the adaptive histogram equalization algorithm to improve its abilities in preserving information entropy, maintaining the mean brightness and retaining the naturalness of the image. The modified technique is known as Entropy-based Adaptive Histogram Equalization Algorithm (EAHEA). With the introduction of a function to determine the value of the adaptive parameter, Beta, EAHEA enables different degrees of enhancement according to the intensity exposure of the input image. The spacing of the two adjacent gray levels of the histogram is adjusted to produce image with improved visual quality. Experiment results show that our technique is excellent in preserving the entropy of the image while the output image appears more natural.