Edge enhancement using adaptive lateral inhibition

Lateral inhibition is a technique to enhance the sharpness of images, which is developed from Hartline´s research on sensory inhibition. Conventional lateral inhibition operates on the entire image like high-pass filter, which also strengthens noise besides edges. Thus this paper proposes an adaptive lateral inhibition based on the features of the image´s color distribution detected by improved Ruzon compass operator. In the proposed algorithm, edge´s strength, orientation and two sides´ extension to blurred edge are detected from color distribution in sub-region. Through given functions with calculated strength, anisotropy and polarization, it is easy to generate corresponding inhibition intensity distribution to enhance the edges of the image. Experimental results show that the proposed method can significantly improve image´s visual effect, whose each edge is enhanced with particular strength, color and orientation while the noise will not.