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

Candidate region extraction method, system and device, medium and target detection method

A technology of candidate area and extraction method, applied in the field of image processing, to achieve the effect of improving speed, high accuracy, and avoiding loss of accuracy

Active Publication Date: 2022-02-08
CHENGDU UNION BIG DATA TECH CO LTD
View PDF18 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0028] The main purpose of the present invention is to solve the problem of the calculation speed of candidate area extraction in the two-stage target detection model

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
  • Candidate region extraction method, system and device, medium and target detection method
  • Candidate region extraction method, system and device, medium and target detection method
  • Candidate region extraction method, system and device, medium and target detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0127] Please refer to figure 1 , figure 1 It is a schematic flow chart of the candidate region extraction method, and the method specifically includes:

[0128] Evenly place several initial candidate regions in the input image;

[0129] Perform several expansion processes on the boundary of the initial candidate area, after the expansion process, correspondingly obtain a first candidate area and a first parameter characterizing the similarity of the first candidate area, each of the initial candidate areas corresponds to Combining multiple first candidate regions to obtain a set of first candidate regions;

[0130] Merging the first candidate region set corresponding to the initial candidate region to obtain a global candidate region set;

[0131] sorting the multiple first candidate regions in the global candidate region set in descending order according to the size corresponding to the first parameter to obtain a first sorting result;

[0132] Selecting the first candid...

Embodiment 2

[0187] Please refer to Figure 4 , Figure 4 It is a schematic diagram of the composition of the candidate region extraction system. Embodiment 4 of the present invention provides a candidate region extraction system, and the system includes:

[0188] An initial candidate area placement unit is used to uniformly place several initial candidate areas in the input image;

[0189] An extension processing unit, configured to perform several times of extension processing on the boundary of the initial candidate area, and obtain a first candidate area and a first parameter characterizing the similarity of the first candidate area after the extension processing, each Combining multiple first candidate regions corresponding to the initial candidate region to obtain a set of first candidate regions;

[0190] a merging unit, configured to merge the first candidate region set corresponding to the initial candidate region to obtain a global candidate region set;

[0191] A sorting unit...

Embodiment 3

[0194] Embodiment 3 of the present invention provides a target detection method, the method comprising:

[0195] Processing the input image by using the candidate region extraction method to obtain a candidate region extraction result of the input image;

[0196] Classify the extracted candidate regions to obtain classification results;

[0197] A target detection result is obtained based on the classification result.

[0198] The target detection method in this embodiment can quickly and accurately realize target detection.

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 candidate area extraction method, system, device, medium, and target detection method, and relates to the field of image processing, including: placing several initial candidate areas in an input image; Corresponding to obtaining a first candidate region and a first parameter characterizing the similarity of the first candidate region, combining multiple first candidate regions corresponding to each initial candidate region to obtain a set of first candidate regions; merging each initial candidate region The corresponding first set of candidate regions obtains a set of global candidate regions; sorting the plurality of first candidate regions in the set of global candidate regions in descending order according to the size corresponding to the first parameter to obtain a first sorting result; selecting the first sorting result The first candidate regions in the top ranks are used as the candidate region extraction results of the input image; this method has high calculation efficiency and detection accuracy.

Description

technical field [0001] The present invention relates to the field of image processing, in particular, to a candidate region extraction method, system, device, medium, and object detection method. Background technique [0002] Object detection is an important recognition task in the field of computer vision, which widely exists in application scenarios such as security monitoring, automatic driving, smart medical care, and smart industry. The target detection algorithm needs to find a specific target from the image, and accurately give the position and size of the target, and then further identify the type of the target. For example, in automatic driving tasks, it is often necessary to detect targets such as vehicles, pedestrians, traffic signs, and obstacles in the vehicle's field of vision, and then analyze the target distance, moving direction, and speed based on the target detection results. Therefore, on the one hand, the target detection algorithm needs to have high ac...

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): G06V10/22G06T7/13G06T7/90
CPCG06T7/13G06T7/90
Inventor 不公告发明人
Owner CHENGDU UNION BIG DATA TECH CO LTD
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