Image Synthesis Applying a Colorimetry-based Color Correction Method

Abstract

A well-known chromatic problem causes that humans may differently recognize the color of the foreground object before and after applying image synthesis. A small number of studies have addressed to solve this problem from the viewpoint of color science. In this paper, we propose a colorimetry-based color correction method for image synthesis. In particular, we apply the Wavelet analysis to calculate the amount of color illusion caused by differences in brightness, saturation, and hue when foreground objects are cut out of the original image and overlaid with another background image. This color correction technique is expected to weaken the illusion of color change when the foreground object in the original image is overlaid with another background image, and to realize more natural image synthesis.