Development of a Level Adaptive Image Enhancer for the Endoscopic Image

抄録

This paper presents a new level adaptive image enhancer for blurred or dark areas in endoscopic images using an adaptive histogram equalization method and an adaptive histogram modification method. In general, conventional simple histogram equalization algorithms can degrade original image quality due to saturation of black or white areas. But the proposed processing can show doctors enhanced image information without blurs in dark areas caused by distance limits that light sources or internal shapes affect. For developing the proposed system, we first defined a classification of histogram patterns in three types for images of blurred or dark areas selected. Then we applied an adaptive histogram equalization method and an adaptive histogram modification method to get a good image quality for degrading areas. Using the proposed level adaptive image enhancing method resulted in more accurate diagnosis.