Variational Specular Separation Using Color and Polarization

Dae Woong Kim, Stephen Lin, Ki-Sang Hong, Heung-Yeung Shum
2002 IAPR International Workshop on Machine Vision Applications  
In this paper, we propose a method that separates specular reflections from polarized color images. With line constraints in RGB color space and spatial domain information, we used a Variational energy minimization approach to restore diffuse reflections. With anisotropic smoothing, our method preserves the shape of textures and successfully separates specular reflections. Also, saturation of specular reflections is considered using a simple inpainting method.
dblp:conf/mva/KimLHS02 fatcat:vk7qbgpcubhqpbprrshkbgq3zm