Local Binary Patterns for Still Images

Abstract

The local binary pattern operator is an image operator which transforms an image into an array or image of integer labels describing small-scale appearance of the image. These labels or their statistics, most commonly the histogram, are then used for further image analysis. The most widely used versions of the operator are designed for monochrome still images but it has been extended also for color (multi channel) images as well as videos and volumetric data. This chapter covers the different versions of the actual LBP operator in spatial domain [42,45,53], while Chap. 3 deals with spatiotemporal LBP [88]. Parts II to IV of this book discuss how the labels are then used in different computer vision tasks. Basic LBPThe basic local binary pattern operator, introduced by Ojala et al. [52], was based on the assumption that texture has locally two complementary aspects, a pattern and its strength. In that work, the LBP was proposed as a two-level version of the texture unit [74] to describe the local textural patterns.The original version of the local binary pattern operator works in a 3 × 3 pixel block of an image. The pixels in this block are thresholded by its center pixel value, multiplied by powers of two and then summed to obtain a label for the center pixel. As the neighborhood consists of 8 pixels, a total of 2 8 = 256 different labels can be obtained depending on the relative gray values of the center and the pixels in the neighborhood. See Fig. 1.1 for an illustration of the basic LBP operator. An example of an LBP image and histogram are shown in Fig. 2.1. Derivation of the Generic LBP OperatorSeveral years after its original publication, the local binary pattern operator was presented in a more generic revised form by Ojala et al. [53]. In contrast to the basic M. Pietikäinen et al., Computer Vision Using Local Binary Patterns, Computational Imaging and Vision 40,

Glasses detection on real images based on robust alignment

“…In experiments with facial images [32], it was found that 90.6 % of the patterns in the (8, 1) neighborhood and 85.2 % of the patterns in the (8, 2) neighborhood are uniform. Using uniform patterns instead of all the possible patterns has produced better recognition results in many applications [35]. By varying sampling radius R, LBP of different resolutions can be obtained, and thus, multiresolution analysis can be accomplished by combining the information provided by multiple operators varying (P, R) [36].…”

Fingerprinting of deformed paper images acquired by scanners

“…They are used in image processing for edge detection [5], texture representation [4] and texture discrimination [6]. Subsequently, they are found to be useful to process images prior to the LBP's feature extraction [12]. A Gabor filter basic formula is represented by , where is the Gabor filter carrier (complex sinusoidal), is the Gabor filter envelope (2D Gaussian-shaped function):…”

Implementation and Evaluation of Face Recognition Based Identification System

“…It thresholds the neighbourhood of each pixel and regards the result as a binary number. Because of its computational simplicity and discriminative power LBP texture operator has become a popular [14].…”