Publication

Abstract : In this chapter, a review of local feature extraction methods and resources for texture recognition is presented, also emphasizing the most dominant approaches employed in these recent years. Here, the performance comparison of several local descriptors, such as local binary pattern (LBP), local directional pattern (LDP), local directional number pattern (LDNP), angular local directional pattern (ALDP), local optimal oriented pattern (LOOP), local line directional neighborhood pattern (LLDNP), volumetric local directional triplet patterns (VLDTP), local tri-directional patterns (LTriDP), local neighborhood intensity pattern (LNIP), and local triangular coded pattern (LTCP), which are commonly used for texture classification, is presented. Experiments are conducted on different texture datasets, such as Brodatz, Outex, Vistex, KTH_TIPS 2a&2b, Kylberg, Describable Texture Datasets, and Salzburg. The findings showed that descriptors such as ALDP, LLDNP, LDRP, and LTCP, with the combination of random forest, give excellent recognition rates of 98.07%, 99.12%, 98.88%, and 98.17%, respectively, on Kylberg dataset.