Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Clothes image classification and retrieval method and system based on convolutional neural network

A convolutional neural network and image technology, applied in the field of image recognition, can solve the problems of low accuracy, low detection efficiency, poor applicability, etc., and achieve the effect of improving accuracy and high robustness

Inactive Publication Date: 2019-09-06
广州瑞智华创信息科技有限公司 +1
View PDF9 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] For this reason, the embodiment of the present invention provides a clothing image classification and retrieval method and system based on a convolutional neural network to solve the problems of low accuracy, low detection efficiency, and poor applicability of existing clothing image retrieval methods

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Clothes image classification and retrieval method and system based on convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0038] like figure 1 As shown, this embodiment proposes a clothing image classification and retrieval method based on a convolutional neural network, which can be used for the classification and retrieval of clothing images. The method includes:

[0039] S100. Divide the clothing image data set with category information into a training set and a verification set.

[0040] In this embodiment, the split ratio of the training set and the verification set is 4:1, and under the premise of ensuring that the images are not distorted, the images of the training set and the verification set are preprocessed to a fixed size of 256×256.

[0041] S200. Construct a convolutional neural network model.

[0042] The convolutional neural network model in this embodiment mainly refers to the resnet network structure idea, and mainly designs a 3×3 convolution kernel. Further, the convolutional neural network model includes sequentially connected first convolutional layer, first ReLU activation...

Embodiment 2

[0056] Corresponding to the above-mentioned embodiment 1, an image classification and retrieval system based on a convolutional neural network is proposed, which includes:

[0057] The data set processing module is used to divide the clothing image data set with category information into a training set and a verification set;

[0058] A model building block for building a convolutional neural network model;

[0059] A model training module for training the convolutional neural network model using the training set;

[0060] Model verification module, for using verification set to verify the convolutional neural network model after training;

[0061] The feature vector library building block is used to extract the feature vector of the image in the clothing image data set using the convolutional neural network model, and establishes the feature vector library;

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The embodiment of the invention discloses a clothes image classification and retrieval method based on a convolutional neural network. The method comprises steps of extracting a feature vector of a to-be-retrieved image through constructing and training a deep convolutional neural network model; performing cosine similarity comparison on the feature vector of the to-be-retrieved image and the feature vector in the established feature vector library, so that several images which are the same as or most similar to the to-be-retrieved image can be retrieved efficiently and accurately. The convolutional neural network has a certain degree of invariance to geometric transformation, deformation and illumination, so that the method can greatly improve the accuracy of image classification and retrieval, is applied to clothing image retrieval, is convenient and efficient, and has relatively high robustness.

Description

technical field [0001] The embodiment of the present invention relates to the technical field of image recognition, and in particular to a clothing image classification and retrieval method and system based on a convolutional neural network. Background technique [0002] With the development of clothing e-commerce, the classification and retrieval technology of clothing images are also gradually updated to meet the changing needs of customers. The traditional image retrieval method is to retrieve clothing images through keywords or texts, and its essence is to search images by text. With the increase of the number of clothing images, the shortcomings of this method become more and more obvious. First of all, keywords can only describe easy-to-extract, abstract semantic features, and cannot fully reflect the visual features of clothing images, especially some fine and difficult-to-describe features; second, due to the huge number of images, it takes a lot of manpower Materia...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06N3/04
CPCG06N3/045G06F18/214G06F18/241
Inventor 张瑞华
Owner 广州瑞智华创信息科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com