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

Complex object automatic recognition method based on multi-category primitive self-learning

A self-learning, complex target technology, applied in character and pattern recognition, instruments, computer components, etc., can solve problems such as automatic recognition of various types of targets

Active Publication Date: 2015-05-06
济钢防务技术有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to provide a complex target automatic recognition method based on autonomous learning of multi-type primitives to solve the problem of automatic recognition of multiple types of targets in images

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
  • Complex object automatic recognition method based on multi-category primitive self-learning
  • Complex object automatic recognition method based on multi-category primitive self-learning
  • Complex object automatic recognition method based on multi-category primitive self-learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0133] An automatic complex target recognition method based on autonomous learning of multiple types of primitives in the present invention proposes a new primitive definition, acquisition and calculation method, integrates three types of primitives, boundary, area and feature point, and uses scale, angle, and target centroid Equal space transformation and constraints independently seek the optimal match, and train strong classifiers separately through independent learning, and realize the positioning, extraction and recognition of multi-type targets in the probability voting space. The present invention overcomes the constraints of image scaling, rotation, side view and other distortions in existing complex target recognition methods, improves training efficiency and recognition performance, improves recognition accuracy and intelligence, and can meet various types of ground object targets recognition and image interpretation needs.

[0134] In the first step, image represent...

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 relates to a complex object automatic recognition method based on multi-category primitive self-learning, which comprises the steps of: a) establishing a representational set of multi-category object images; b) preprocessing images in a training set and respectively extracting point, linear and planar primitives; c) conducting concentrated matching calculation, screening and merging to the obtained numerous primitives in a confirmation image set, and respectively constructing point, linear and planar primitive dictionaries; and d) selecting a certain quantity of primitives from the dictionaries, using the primitives as a weak classifier after the primitives are mated and combined, and respectively training the strong classifiers of the three categories of primitives through self-learning; and e) combining the strong classifiers of the three categories of primitives in a probabilistic polling space to realize the accurate positioning, contour extraction and categorical recognition of multi-category complex objects. The method provided by the invention has the advantages that the intelligent level is high and the demands for the recognition and image interpretation of multi-category complex objects can be met.

Description

technical field [0001] The invention relates to a method for target recognition in the technical field of image processing, and is a method for realizing automatic recognition of multiple types of complex targets in an image by synthesizing multiple types of primitives and through autonomous learning. Background technique [0002] With the advancement of information storage and transmission technology, the number of images has shown an explosive growth, and the application field of image processing technology has also continued to expand. The traditional method of recognizing and interpreting objects in images by manual labor has become infeasible due to the large amount of manpower and material resources required. Therefore, as an important link in image processing technology, automatic target recognition has increasingly become the basis of various image processing applications. Due to the large amount of data in the existing images, various types of targets, complex stru...

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): G06K9/66
Inventor 孙显付琨王宏琦
Owner 济钢防务技术有限公司
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