Image Enhancement for Dichromats by Weighting the Color Difference Between Neighboring Pixels

Dichromats recognize colors using two out of three cone cells; L, M, and S. To extend the ability of dichromats to recognize the color difference, we propose a method to expand the color difference when observed by dichromats. We analyze the color between the neighboring pixels in chromaticity space. In order to exaggerate the neighboring pixels as much as possible, we set the initial value randomly for iterative updation. On the other hand, in order to preserve the naturalness of the detailed structure, we weight parameters adequately to the cost function. Our method fully enhances the color difference between different color region so that the dichromats recognizes the color difference between neighboring pixels.