A method for extracting target features, a target feature extraction module, a target model creation module, and an intelligent image monitoring device

A target feature and target technology, applied in image analysis, image data processing, character and pattern recognition, etc., can solve the problems of low target tracking reliability, mistracking target, easy loss, etc., to improve target tracking efficiency, improve The effect of extracting quality and reducing the amount of computation

Active Publication Date: 2019-06-11
上海圣尧智能科技有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The problem solved by the present invention is that in the prior art, the target feature extraction method mainly adopts the method of non-overlapping segmentation target discovery frame for color histogram extraction, so that the reliability of target tracking is not high, and it is easy to lose or mistakenly track the target

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 method for extracting target features, a target feature extraction module, a target model creation module, and an intelligent image monitoring device
  • A method for extracting target features, a target feature extraction module, a target model creation module, and an intelligent image monitoring device
  • A method for extracting target features, a target feature extraction module, a target model creation module, and an intelligent image monitoring device

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0052] refer to figure 2 , the present embodiment provides a method for target feature extraction, including:

[0053] In the first step, the target discovery box locks the target class.

[0054] Before the monitoring system starts shooting, a large amount of target class contour information, that is, the direction gradient histogram, is stored in the target discovery module. If the matching degree is greater than the threshold, it is considered as the target class that needs to be tracked.

[0055] After the monitoring system discovers the target class, it locks the targets in the target class through the target discovery frame, so that the target is surrounded by the target discovery frame. The target detection frame is generally a rectangular frame, and the target class can be a person or other moving objects. If there are multiple targets in the target class, multiple target discovery boxes can be created so that each target discovery box contains a target.

[0056] T...

no. 2 example

[0078] The second embodiment of the present invention provides an object feature extraction module formed by the method of the first embodiment.

[0079] Compared with the prior art, the technical solution of the present invention has the following advantages:

[0080] The target discovery frame is divided into several sub-frames that overlap locally, and the target contour area can be contained by more sub-frames, that is to say, more sub-frames can contain the contour area. Since the texture features of the color of the outline area of ​​the target are more abundant, more sub-frames are included in this color-textured area, and the color histogram of the sub-frame can better reflect the characteristics of the target, which is conducive to improving the reliability of target tracking.

no. 3 example

[0082] The third embodiment of the present invention provides an object model creation module formed by the method of the first embodiment.

[0083] Compared with the prior art, the technical solution of the present invention has the following advantages:

[0084] The target discovery frame is divided into several sub-frames that overlap locally, and the target contour area can be contained by more sub-frames, that is to say, more sub-frames can contain the contour area. Since the texture features of the color of the outline area of ​​the target are more abundant, more sub-frames are included in this color-textured area, and the color histogram of the sub-frame can better reflect the characteristics of the target, which is conducive to improving the reliability of target tracking.

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

A method for object feature extraction, comprising: the first step, the target detection frame is locked to the target class; the second step, the target detection frame is divided, so that the target detection frame is divided into several sub-frames that overlap locally; the third step is to calculate each sub-frame The color histogram in . The target discovery frame is divided into several sub-frames that overlap locally, and the target contour area can be contained by more sub-frames, that is to say, more sub-frames can contain the contour area. Since the texture features of the color of the outline area of ​​the target are richer, more sub-frames are included in this color-textured area, and the color histogram of the sub-frame can better reflect the characteristics of the target, which is conducive to improving the reliability of target tracking.

Description

technical field [0001] The invention relates to the field of intelligent image monitoring, in particular to a method for extracting object features, an object feature extraction module, an object model creation module and an intelligent image monitoring device. Background technique [0002] With the advent of the Internet of Things era, smart products continue to penetrate into our daily life, and put forward higher requirements for the performance of previous smart products. [0003] Taking intelligent image monitoring as an example, it mainly includes a target discovery module, a target model creation module and a target tracking module. [0004] The target discovery module is used to locate possible target classes. After the target class is correctly located, the target model creation module is used to extract the features of the target of interest. After the target features are extracted, the target tracking module is used to track the target. [0005] The current monit...

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): G06K9/46G06T7/246
CPCG06V10/56G06V10/50
Inventor 张杰
Owner 上海圣尧智能科技有限公司
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