Improved Blur Insensitivity for Decorrelated Local Phase Quantization

This paper presents a novel blur tolerant decorrelation scheme for local phase quantization (LPQ) texture descriptor. As opposed to previous methods, the introduced model can be applied with virtually any kind blur regardless of the point spread function. The new technique takes also into account the changes in the image characteristics originating from the blur itself. The implementation does not suffer from multiple solutions like the decorrelation in original LPQ, but still retains the same run-time computational complexity. The texture classification experiments illustrate considerable improvements in the performance of LPQ descriptors in the case of blurred images and show only negligible loss of accuracy with sharp images.

• Type: Conference
• Year: 2010
• Where: ICPR
• Authors: Janne Heikkilä, Ville Ojansivu, Esa Rahtu
• Keywords: Computer Vision, ICPR 2010, Kind Blur, Local Phase Quantization, Texture Descriptor