Chromaticity improvement using the MSR model in presence of shadows

Abstract

One of the main problems in digital images is the illumination conditions affecting different objects due to angle reflection and lightness in a scenario, to solve this the Retinex algorithm is proposed to estimate the illumination source due to the lighting conditions, which in turn causes that the artificial vision algorithms deliver little optimum visual results; particularly this phenomenon is caused by the shadows and the angles of the incident source that produces different reflections from different objects in a scene. So, in this article an algorithm is proposed to diminish the effect of the shadows present in the digital images using the method of Color Constancy of a pixel. The actual proposal presents good inherent characteristics preservation, such as contrast and poor visibility.