Abstract
Texture is an important cue for different computer vision tasks andapplications. Local Binary Pattern (LBP) is considered one of the best yetefficient texture descriptors. However, LBP has some notable limitations,mostly the sensitivity to noise. In this paper, we address these criteria byintroducing a novel texture descriptor, Robust Adaptive Median Binary Pattern(RAMBP). RAMBP based on classification process of noisy pixels, adaptiveanalysis window, scale analysis and image regions median comparison. Theproposed method handles images with high noisy textures, and increases thediscriminative properties by capturing microstructure and macrostructuretexture information. The proposed method has been evaluated on popular texturedatasets for classification and retrieval tasks, and under different high noiseconditions. Without any train or prior knowledge of noise type, RAMBP achievedthe best classification compared to state-of-the-art techniques. It scored morethan $90\%$ under $50\%$ impulse noise densities, more than $95\%$ underGaussian noised textures with standard deviation $\sigma = 5$, and more than$99\%$ under Gaussian blurred textures with standard deviation $\sigma = 1.25$.The proposed method yielded competitive results and high performance as one ofthe best descriptors in noise-free texture classification. Furthermore, RAMBPshowed also high performance for the problem of noisy texture retrievalproviding high scores of recall and precision measures for textures with highlevels of noise.