Previous [ 1] [ 2] [ 3] [ 4] [ 5] [ 6] [ 7] [ 8] [ 9] [ 10] [ 11] [ 12] [ 13] [ 14] [ 15]


Journal of Information Science and Engineering, Vol. 29 No. 5, pp. 835-849 (September 2013)

Specularity Removal using Dark Channel Prior*

School of Information Science and Engineering
Central South University
Changsha, 410083 P.R. China

The reflectance of inhomogeneous objects can be described as a linear combination of diffuse and specular reflection components. Most computer vision algorithms assume that visually observable surfaces consist only of diffuse reflection. The existence of specular reflection can be misleading to these computer vision algorithms. A new algorithm { dark channel prior based specularity removal is proposed for separating specular and diffuse reflection components on colorful surfaces from a single input image. The dark channel prior is applied to detect the specular pixels in the image. The maximum diffuse chromaticity of the diffuse pixels is propagated to their neighboring specular pixels after specularity have been detected. Specularity removal can be achieved by using the specular-to-diffuse mechanism. The experimental results show that the proposed algorithm obtain comparable results as the state-of-the-art reflection components separation methods with the merit of being computationally more efficient.

Keywords: specularity removal, specularity detection, chromaticity, dark channel prior, specular-to-diffuse mechanism

Full Text () Retrieve PDF document (201309_03.pdf)

Received December 9, 2011; revised March 12, 2012; accepted May 20, 2012.
Communicated by Yung-Yu Chuang.
* This work was supported by the National Natural Science Funds of China (No. 60970098 and No. 61173122), the Doctoral Program Foundation of Institutions of Higher Education of China (No.20090162110055, No. 200805331107), Hunan Provincial Natural Science Foundation of China (No. 09JJ6102).