Local binary pattern texture image feature extraction method and system based on scale and angle adaptive selection

Abstract

The invention discloses a local binary pattern texture image feature extraction method and system based on scale and angle adaptive selection, and the method comprises the steps: obtaining a texture image of a to-be-extracted feature, and obtaining gradient values of each central pixel in eight directions through employing a Kirsch gradient operator; taking the gradient value with the maximum absolute value as the gradient value for measuring the change speed of the local area where each central pixel is located; equally dividing and classifying the absolute values of the gradient values of all the central pixels in eight directions to obtain a classification result; obtaining final neighborhood pixels of each central pixel in eight directions; and generating a joint feature histogram according to the symbol information of the difference vector of each central pixel in the whole texture image, the amplitude information of the difference vector and the central pixel information. According to the invention, the robustness of the local binary pattern algorithm to rotation, illumination and scale change can be enhanced; classification accuracy can be improved.

Description

technical field [0001] The invention belongs to the technical field of texture image feature extraction and processing, and in particular relates to a local binary mode texture image feature extraction method and system based on adaptive selection of scale and angle. Background technique [0002] Texture is composed of many periodically repeated, close to each other and intertwined structures. As an inherent property of the surface of an object, it has three major characteristics: local sequence changes periodically, orderly arrangement, and uniform uniform distribution within a local range. Different from the color and edge features of the object, the local texture features describe the distribution characteristics of the pixel and its surrounding neighborhood space, while the global feature is the regular repeated composition of the local texture features. Human beings can extract distinguishable features such as color, edge and texture features of objects through their ow...

AI Technical Summary

Problems solved by technology

However, the existing LBP algorithm and its many improved algorithms still have limitations, including: (1) limited by the fixed radius neighborhood, unable to obtain multi-scale features; (2) sensitive to illumination and rotation changes, seriously affecting The final classification performance of the algorithm

Images