Clothes classification and identification method based on Weber local descriptor

Abstract

The invention discloses a clothes classification and identification method based on a Weber local descriptor, and belongs to the technical fields of image processing and model identification. The clothes classification and identification method comprises the following steps: performing feature extraction on a clothes image serving as a training sample and a clothes image to be classified by using a Weber local descriptor to obtain a feature vector for representing the clothes image; evaluating the similarity between feature vector of the clothes image to be classified and the feature vector of the clothes image in the training sample; selecting clothes to be classified which have highest similarity with the training sample, and selecting clothes to be classified which have highest similarity with the training sample through the difference between the two vectors. Through adoption of the clothes classification and identification method, the multi-classification problem can be solved by using a small quantity of training samples, and meanwhile the quantity of training samples needing to be provided by the each kind of clothes is small; moreover, clothes styles can be described quantitatively, and the information of clothes can be inquired rapidly and conveniently.

Description

technical field [0001] The invention discloses a clothing classification recognition method based on Weber local descriptor, which belongs to image processing and pattern recognition technology. Background technique [0002] At present, the commonly used methods for item classification and recognition mainly include pattern classification methods based on SVM (support vector machine), pattern classification methods based on BP artificial neural network, and pattern classification methods based on Adaboost algorithm. But these methods have their own limitations. [0003] The pattern classification method of SVM (Support Vector Machine) is to map the sample space to a high-dimensional or even infinite-dimensional feature space (Hilbert space) through a nonlinear mapping p, so that the non-linearly separable The problem is transformed into a linearly separable problem in the feature space, but SVM (Support Vector Machine) is more effective for binary classification problems, a...

AI Technical Summary

Problems solved by technology

If there are too few training samples, better classification results cannot be obtained

Images