Application of Adaptive Image Enhancement Method for Mammogram Enhancement

Enhancing mammograms plays a critical role in improving contrast, thereby aiding radiologists in the accurate diagnosis of breast cancer. This study introduces a novel approach that combines low illumination techniques, a color image correction approach grounded in nonlinear functional transformations based on the illumination-reflection model, and multiscale theory to effectively enhance mammographic contrast. Initially, the 2D mammogram is transformed into a 3D representation for enhancement and subsequently transformed to the HSV color model. The illumination component of the V channel is extracted using multiscale Gaussian functions. Adaptive adjustments are then applied to the image improvement function parameters for both S and V channels based on the distribution profiles of the illumination elements. Furthermore, an image fusion method is devised to extract details from both images, followed by conversion from HSV to RGB space. Finally, the RGB image is complemented to obtain an enhanced image. Experimental findings demonstrate that the suggested approach significantly enhances mammographic contrast while mitigating oversaturation issues observed in bright regions when compared to existing methods.