Detecting Local Illumination Using Skewness of Oriented Gradients from a Single Image
In this paper we present a simple and effective method for detecting illumination of a region from a single image. Our method is primarily based on skewness, which is a measure of asymmetry of a data set in statistics. We happen to find out that the skewness value of oriented gradients of an image can measure the directional characteristic of illumination. By choosing appropriate statistical area, we can analyze the subtle changes on the surface of an object. Theoretical analysis and experimental results show that our algorithm is accurate and effective. In the end, we give its application in image authenticity verification problem which is to distinguish real and “flat” objects in a photograph, and it shows excellent results.
F. Zhang et al., "Detecting Local Illumination Using Skewness of Oriented Gradients from a Single Image", Applied Mechanics and Materials, Vols. 58-60, pp. 2381-2386, 2011