A spatially controlled histogram equalization for image enhancement

Image enhancement mostly means handling with the contrast of the pixels. However, conventional histogram equalization (HE) methods works based on the global histogram information only. Hence they often make an overall enhancement of the whole image without taking any intensive care of the spatial relationship among the pixels. This often leads to a number of annoying artifacts. In this paper we propose to incorporate spatial information in histogram equalization process to apply the appropriate amount of contrast enhancement for a visually pleasing output. This method outperforms other present HE approaches by enhancing the contrast well without any severe side affects such as washed out appearance (caused by over-enhancing some of the features and under-enhancing some others), checkerboard effects etc.