Multi-resolution target recognition method and device

A target recognition and multi-resolution technology, which is applied in the field of multi-resolution target recognition methods and devices, can solve the problems of increased difficulty in identifying targets, decreased accuracy, and large image memory usage, achieving a balance between recognition efficiency and memory usage , Improve efficiency and scope of application

Active Publication Date: 2022-03-29
HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] The target recognition in the screen through the real-time monitoring screen has always been the research focus of computer vision, and it has applications in various fields such as security, monitoring, access control, etc. Usually, the image is collected in real time through the camera and sent to the processing system for target detection. And recognition, but if the resolution of the camera is low, the clarity of the image it collects is low, the difficulty of identifying the target from the image increases and the accuracy decreases. If a high-resolution camera is used, the captured image is clear and easy to identify the target , but the image captured by the high-resolution camera takes up a lot of memory and increases the processing time

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
  • Multi-resolution target recognition method and device
  • Multi-resolution target recognition method and device
  • Multi-resolution target recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0045] Embodiment 1 of the present invention provides a multi-resolution target recognition method, figure 1 An implementation flow chart of a multi-resolution target recognition method provided by an embodiment of the present invention, as shown in figure 1 As shown, the method includes the following steps:

[0046] S1: Obtain the first image of the first resolution of the area to be tested, specifically to monitor a larger field of view at a lower resolution. It can be understood that in this embodiment, different resolutions are achieved through a camera device capable of automatic zooming image acquisition

[0047] S2: Use the target detection model to perform the first target detection in the first image, obtain the first identification result of the area to be tested, and obtain the confidence of the first identification result, and realize the preliminary identification in the low-resolution first image. There are targets to be tested, and the targets to be tested ref...

Embodiment 2

[0082] This embodiment provides a multi-resolution target recognition device for performing the method described in Embodiment 1, such as image 3 As shown, it is a structural block diagram of the multi-resolution target recognition device of this embodiment, including:

[0083] Obtaining the first image module 10: used to acquire the first image of the first resolution of the area to be tested;

[0084] The first target detection module 20: using the target detection model to perform the first target detection in the first image, obtain the first recognition result of the area to be tested, and obtain the confidence of the first recognition result;

[0085] The first result output module 30: used to directly output the first identification result when the confidence level is greater than the preset confidence level;

[0086] The second target detection module 40: used to take the area to be tested as the center area, acquire a second image with a second resolution in a prese...

Embodiment 3

[0088] This embodiment provides a device for identifying a human object, using the method described in Embodiment 1 to identify a human object, the human object includes: a human face or a human posture, that is, in this embodiment, the target to be detected is a human face or a human posture, It can be understood that the training set includes a large number of image samples of faces or poses of people for training the object recognition model.

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 multi-resolution target recognition method and device. It relates to the field of target recognition, wherein, the method acquires the first image of the first resolution of the area to be tested, uses the target detection model to perform the first target detection, and obtains the first recognition result and the corresponding confidence degree. When the confidence degree is greater than the preset confidence directly output the first recognition result, otherwise, obtain the second image with the second resolution in the preset range, use the target detection model to perform the second target detection in the second image, obtain the second recognition result and output it, where the second resolution The rate is greater than the first resolution, which realizes a large-scale monitoring through low resolution. When the target to be measured appears, the resolution is increased to identify and confirm the target to be measured. Through the state switching of the two resolutions, the recognition efficiency and occupancy are balanced. Memory improves the efficiency and scope of application of target recognition and detection, and can be widely used in target recognition and related fields.

Description

technical field [0001] The invention relates to the field of target detection, in particular to a multi-resolution target recognition method and device. Background technique [0002] The target recognition in the screen through the real-time monitoring screen has always been the focus of computer vision research, and it has applications in various fields such as security, monitoring, access control, etc. Usually, the image is collected in real time through the camera and sent to the processing system for target detection. And recognition, but if the resolution of the camera is low, the clarity of the image it collects is low, the difficulty of identifying the target from the image increases and the accuracy decreases. If a high-resolution camera is used, the collected image is clear and easy to identify the target , but the image captured by the high-resolution camera takes up a lot of memory and increases the processing time. Therefore, it is necessary to propose a target ...

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): G06V10/776G06V10/82G06V10/774G06V20/52G06V40/10G06V40/16G06K9/62
CPCG06V40/161G06V20/53G06V20/584G06V20/52G06F18/22
Inventor 徐勇王俊茜
Owner HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL
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