Image location marking and identifying method based on view overlapping

An image and field of view technology, applied in the field of image location labeling and recognition based on overlapping fields of view, can solve the problems of difficult model training, inability to accurately obtain positive samples of query images, etc., and achieve the effect of high recall rate

Pending Publication Date: 2022-07-22
INST OF INFORMATION ENG CHINESE ACAD OF SCI
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, only using GPS annotations cannot accurately obtain positive samples of query images, which brings difficulties to model training

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
  • Image location marking and identifying method based on view overlapping
  • Image location marking and identifying method based on view overlapping
  • Image location marking and identifying method based on view overlapping

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] In order to make the above-mentioned features and advantages of the present invention more obvious and easy to understand, the following embodiments are given and described in detail with the accompanying drawings as follows.

[0042] The embodiment of the present invention proposes an image location labeling and recognition method based on overlapping fields of view. First, multi-threshold screening is used to mine potential positive samples of the query image, then the multi-scale sub-regions of the image are cropped and the similarity between the image regions is calculated. The regional similarity filters the pseudo-positive samples and identifies the overlapping regions of the field of view between the query image and the real positive samples, and constructs field-of-view overlapping annotations for the dataset. Further use the dataset with overlapping visual field annotations to train the model, use NetVLAD to extract the features of images or regions, and calcula...

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 an image location labeling and recognition method based on view overlapping, belongs to the technical field of crossing of computer vision and robots, and aims to construct richer labeling information for an existing image location recognition data set and directly obtain a positive sample of a query image through the labeling. And a model is trained by using the more difficult positive samples, so that more robust image features can be learned.

Description

technical field [0001] The invention belongs to the cross technical field of computer vision and robots, and in particular relates to an image location labeling and identification method based on overlapping visual fields, which can identify the location of a query image in an image database. Background technique [0002] Visual Place Recognition (VPR) is one of the important tasks in the field of computer vision and robotics. It is widely used in technologies such as real-time localization and map construction, virtual reality, and autonomous driving. Its goal is to identify the visited places. The location of the query image is obtained, so as to further predict the pose of the robot or moving body based on visual navigation. The visited location is usually represented as a location-annotated image database. During recognition, an image related to the query image is found in the database, and its location is estimated as the location of the query image. Therefore, image l...

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 Applications(China)
IPC IPC(8): G06V20/10G06V10/20G06V10/40G06V10/74G06V10/82G06V10/77G06N3/04G06N3/08G06F16/535G06F16/29
CPCG06N3/08G06F16/535G06F16/29G06N3/045G06F18/2135G06F18/22
Inventor 王蕊祁俊昆
Owner INST OF INFORMATION ENG CHINESE ACAD OF SCI
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