Increasing Contrast in X-ray Images Using Retinex- and CLAHE-Based Region Segmentation

An X-ray image created with a flat panel detector (digital) in medical imaging often has low image quality, making it difficult to diagnose a patient successfully. The goal of image enhancement techniques is to make the information contained in the images easier to understand. Poor quality and low contrast are the major issues in medical images. Thus, noise reduction and contrast enhancement have become the two major goals of image enhancement research. In this research, the Otsu strategy was used in the segmentation step. Brightness was improved using the Retinex method, whereas contrast was enhanced using the contrast-limited adaptive histogram equation. Four quantitative metrics, i.e. entropy, average gradient, mean of the local standard deviation and contrast enhancement measure, were used to evaluate the enhanced X-ray images. Results showed that the suggested method obtained the best averages for entropy (7.020), average gradient (6.020), mean of the local standard deviation (35.5001) and contrast enhancement measure (0.674).