Avaliação de técnicas de realce de imagens digitais utilizando métricas subjetivas e objetivas

Abstract

Image enhancement is a technique that increases the visual information available in images. Its use is important for many areas of science, engineering, biomedical, among others. The objetive of this thesis is to implement and evaluate five methods of enhancement images based on Histogram Equalization, Gamma Correction and Retinex. In literature there are objective and subjective metrics that help in the extraction of important features, from the enhanced image, thus being able to assist in the diagnosis of their quality. The images used in this thesis are provided by NASA, which also provides its enhancement results which are evaluated objectively. The results of these five methods are compared using three objective metric Entropy, AMBE (Absolute Mean Brightness Error) and EME (Measure of Enhancement); and subjective by visual analysis issued by 35 volunteers. Two methods have demonstrated sufficient performance for use on platforms with low computing power as embedded platforms.