Analyzing contrast enhancement and watershed segmentation using quantitative information measuring

Image enhancement and image segmentation are both among most practical approaches in image processing. There are a wide variety of applications on medical diagnosis, homeland security, weather forecasting and environment protection. Two typical technologies out of these two categories are contrast enhancement and watershed segmentation. Under improper illumination and unexpected disturbances, adaptive image enhancement can be conducted, which adapts to the intensity distribution within that image. For the trimulus true color system, each of the three color components takes an independent role. Watershed segmentation is an effective approach to automatically separate different objects, where watershed lines separate catchment basins. Erosion and dilation operations are implemented in the segmentation process. Foreground and background markers can be applied to avoid over-segmentation. To represent effects of image enhancement and image segmentation, quantitative measures should be taken into account together with the qualitative observations. Based on gray level quantity measures and the expanded true color component quantity measures, component energy, discrete entropy and relative entropy are calculated to indicate the impact of adaptive image contrast enhancement technique and watershed segmentation techniques. Thus useful results on image processing are obtained.