Compensation for specular illumination in acne patients images using median filter

Specular illumination in digital images is one of the main causes, affecting the performance of image analysis techniques. The causes of this artifact include shiny nature of the surface on which the light is falling and the intensity and distribution of light. In this paper, it is explained that how specularly illuminated spots in the color images of acne patients can be detected and compensated for that. For detecting the specular illumination spots, a fixed threshold value were used in each of the three channels of the RGB color space and after that median filtering was performed in G and B channels only. Experimentation was carried out on twenty images of acne patients and subjectively satisfactory results were obtained. The effect of filter size on processing time was also examined and it was found that with increasing filter size, the time complexity of image processing also increases.