A Feature Extraction Method for Two-Dimensional Images

Abstract

The invention discloses a feature extraction method of a two-dimensional image. The method includes the following steps: S1, obtaining Gabor filter differential decomposition for a texture image, and obtaining a set of binary images; S2, using statistical calculation for the decomposed binary image Sub-extracting texture features; S3, using principal component analysis to reduce dimensionality of the extracted texture feature vectors, and performing texture classification and retrieval on the feature vectors after dimensionality reduction. The present invention has the advantage of not needing to define texture primitives, and the calculation complexity is moderate. The newly proposed five statistical operators more comprehensively describe the decomposed texture primitives and the arrangement rules of primitives, and in the recognition process , the introduction of the principal component analysis algorithm increases the calculation speed of the method and is applicable to the real-time retrieval and classification of texture images.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a feature extraction method of a two-dimensional image. Background technique [0002] Texture is an important visual attribute of an image, which is easy to identify but difficult to define. After decades of research, researchers have proposed hundreds of texture description methods. These methods can be roughly divided into statistical methods, structural methods, model methods and signal processing methods. Among them, the structural method is most in line with the cognitive characteristics of human beings. This kind of method believes that the texture image is composed of different texture primitives according to different arrangement rules. For example, in the beach texture, each grain of sand is the texture primitive of the texture, and the random distribution rules among the sands are the arrangement rules of the primitives. [0003] Structural methods were prop...

AI Technical Summary

Problems solved by technology

For the statistical geometric feature method, the global threshold is used to decompose the image, so that the method does not have illumination invariance

Secondly, for the decomposed binary image, only the geometric information of the individual binary patches is extracted, and the distribution information between the patches is not described at all, so that the final description features are incomplete and the recognition ability is limited.

For the statistical multi-scale patch feature method, although the two-dimensional decomposition of the spatial scale and the gray scale improves the description ability of the feature, the dimension of the extracted feature is also greatly increased, which increases the computational complexity.

Images