A Big Data Image Classification Method

A classification method and big data technology, applied in the direction of electrical digital data processing, special data processing applications, instruments, etc., can solve the problems of large number of classified samples, increased development cost of image classification software, high use cost, etc.

Active Publication Date: 2016-10-05
SOUTH CHINA UNIV OF TECH
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, big data image classification faces difficulties such as the huge number of categories and the huge number of samples that need to be classified.
Linear discriminant analysis is relatively expensive for big data. In order to obtain a certain classification performance, it requires a large number of manually labeled samples.
This greatly increases the development cost of image classification software and requires a large number of manually labeled samples

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
  • A Big Data Image Classification Method
  • A Big Data Image Classification Method
  • A Big Data Image Classification Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0047] In order to clearly illustrate the effectiveness of the present invention for image classification, such as figure 1 As shown, in this embodiment, a handwritten digital image classification test is carried out, and compared with the classic linear discriminant analysis (LDA). The test data selects the common USPS data set, the data is from 0 to 9, a total of 10 categories, 9298 samples, the specific implementation steps are as follows: (combining the embodiment and figure 1 Combined to specifically describe the test steps and list the test results):

[0048] Step 1: Collect 10 image samples for each category, a total of 100 samples as the training set X, ie X=[x 1 ,x 2 ,...,x N ]∈R D×N , the sample dimension is D=256, and each sample has a corresponding category mark C i ∈ Z n . The remaining samples are used as the test data set Xu.

[0049] 2) Establish a local optimization objective function:

[0050] For each labeled sample x i , we can find the in-class s...

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 invention discloses a big data image classification method, comprising the following steps: 1) collecting image samples as a training set; 2) finding the optimal projection matrix for big data image classification; 3) projecting unlabeled data; 4) The projected samples are classified using a minimum distance classifier. The method proposed by the invention can effectively utilize the local geometric information of sample distribution, and extract the identification information of classification, reduce the dependence of big data image classification on manually labeled samples, effectively reduce the storage cost in the training process, and its classification accuracy is higher than Representative image classification methods based on linear discriminant analysis.

Description

technical field [0001] The invention relates to an image classification technology in the technical field of pattern recognition and artificial intelligence, in particular to a big data image classification method, which is a supervised learning image classification method. Background technique [0002] With the rapid development of the mobile Internet, more and more smart phones and tablet computers with digital cameras have entered people's lives, and it is easy to generate a large number of personal digital images. Although it is a common method to manage images by using time and directory, it lacks the effective management of images at the semantic level. Therefore, the supervised learning method is used to obtain the image classification model by learning the manually labeled data, and then automatically classify the unlabeled images. Since images usually have very high feature dimensions, dimensionality reduction methods can help improve recognition performance. [0...

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 Patents(China)
IPC IPC(8): G06F17/30G06K9/62
CPCG06F16/583G06F18/24
Inventor 金连文陶大鹏王永飞
Owner SOUTH CHINA UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap