ARTICLE IN PRESS JID: CAEE [m3Gsc;September 7, 2017;21:55] Computers and Electrical Engineering 000 (2017) 1–13 Contents lists available at ScienceDirect Computers and Electrical Engineering journal homepage: www.elsevier.com/locate/compeleceng Texture features based on an efficient local binary pattern descriptor  Bachir Kaddar a,∗ , Hadria Fizazi a , Abdel-Ouahab Boudraa b a Department of Computer Sciences, Université des Sciences et de la Technologie d’Oran Mohamed Boudiaf, USTO-MB, BP 1505, El Mnaouer, 31000 Oran, Algeria b Ecole Navale/Arts & Metiers ParisTech, France a r t i c l e i n f o Article history: Received 10 August 2016 Revised 10 August 2017 Accepted 15 August 2017 Available online xxx Keywords: Texture discrimination Multi-scale representation Bilateral filter Keypoints extraction Scale invariant feature transform Mixed pixels a b s t r a c t Texture characterization aims at describing the spatial arrangement of local structures within an image. However, mixed pixels that are generally located near boundaries of the regions represent challenge to perform accurate image texture discrimination. To address this problem, this paper proposes a robust discriminating texture features relying on an efficient Local Binary Pattern (LBP) descriptor, where the spatial information within image is taken into account. To determine for each pixel both a proper scale parameter and a threshold value to compute the LBP code, an efficient way relying on bilateral filter-based multi-scale image analysis is used. First, the difference of Gaussian operator is used to determine the corresponding scale. Second, key points based-approach is used to identify the threshold value of each pixel. This provides the ability to deal with mixed pixels. Then, LBP code is computed to characterize the texture information for each pixel. Experimental results, using both synthetic and real images, show that the proposed appropriate-scale- threshold selection strategy demonstrates a significant improvement in texture discrimi- nation ability. © 2017 Elsevier Ltd. All rights reserved. 1. Introduction Texture is one of the most important characteristics used in image analysis. It represents the different local structural features and their spatial arrangement within the image [1]. Mainly, because of their robustness against noise and illumina- tion variations, texture features remain almost unchanged and this allows an accurate description of the image, conversely to the spectral intensity. Accordingly, texture feature extraction is an important task for many computer vision applications, such as image change detection and segmentation [2]. However, the challenge is the ability to discriminate of textures rep- resenting different spatial local structures. To cope with this problem, different approaches have been developed [3]. Among them, the grey-level co-occurrence matrix, which has been the focus of interest of an increased number of image processing works [4]. However, this approach is limited by the high computational cost. Local Binary Pattern (LBP) operator is consid- ered as a powerful texture primitive descriptor characterized by its computation simplicity, and few parameters are required to be set [5]. This operator labels the pixels of an image by thresholding the vicinity of each pixel and considers the result as a binary number. One limitation of this operator lies in the fact that the associated parameters are fixed independently  Reviews processed and recommended for publication to the Editor-in-Chief by Guest Editor Dr. A. H. Mazinan. ∗ Corresponding author. E-mail addresses: kaddarbachir@gmail.com, bachir.kaddar@univ-usto.dz (B. Kaddar). http://dx.doi.org/10.1016/j.compeleceng.2017.08.009 0045-7906/© 2017 Elsevier Ltd. All rights reserved. Please cite this article as: B. Kaddar et al., Texture features based on an efficient local binary pattern descriptor, Computers and Electrical Engineering (2017), http://dx.doi.org/10.1016/j.compeleceng.2017.08.009