FWLBP: A Scale Invariant Descriptor for Texture Classification

In this paper we propose a novel texture recognition feature called Fractal Weighted Local Binary Pattern (FWLBP). It has been observed that fractal dimension (FD) measure is relatively invariant to scale-changes, and presents a good correlation with human perception of surface roughness. We have utilized this property to construct a scale-invariant descriptor. We have sampled the input image using an augmented form of the local binary pattern (LBP), and then used an indexing operation to assign FD weights to the collected samples. The final histogram of the descriptor has its features calculated using LBP, and its weights computed from the FD image. The proposed descriptor is scale, rotation and reflection invariant, and is also partially tolerant to noise and illumination changes. In addition, it is also shown that the local fractal dimension is relatively insensitive to the bi-Lipschitz transformations, whereas its extension is able to correctly discriminate between fundamental texture primitives. Experimental results show the proposed descriptor has better classification rates compared to the state-of-the-art descriptors on standard texture databases.