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

Object recognition method and device

An object recognition and object technology, applied in the field of image recognition, can solve problems such as cumbersome recognition steps, changes in national administrative divisions, and low recognition rates

Pending Publication Date: 2021-06-08
CHINA CONSTRUCTION BANK
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The text recognition in the first step includes two steps of text positioning and text recognition, and text positioning has always been a difficult point in OCR recognition (optical character recognition), and it is easy to locate inaccurately, and this text recognition belongs to handwriting recognition, which is different from printing Body recognition, due to irregular writing and differences in personal writing habits, there will be a problem of low recognition rate
The string fuzzy matching in the second step is a method to calculate the similarity of sentences, including edit distance algorithm and fuzzy query algorithm, etc. It usually uses the similarity relationship between words for matching. The biggest problem is that the country’s administrative divisions often change, and Place name changes are not as similar as synonym replacement. For example, "Xuanwu District, Beijing" is changed to "Xicheng District, Beijing". There is no similarity relationship between Xuanwu and Xicheng, so the matching accuracy is low
[0005] Due to the diversity of object content, the recognition accuracy is not high, and the recognition steps are cumbersome

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
  • Object recognition method and device
  • Object recognition method and device
  • Object recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0057] There is one positioning mark, and the positioning mark frame corresponds to the positioning mark one by one, that is, the position of the positioning mark frame in the template image is the same as the position of the positioning mark in the given image, and the position of the positioning mark frame and the slice frame in the template image And the position of the positioning mark in the given image is known, and the relative position between the positioning mark and the object in the image is fixed, then the distance between the positioning mark frame and the slice frame in the template image, and between the positioning mark and the slice frame in the given image The regions corresponding to the objects to be recognized have the same positional relationship. The position of the slice image in the given image can be calculated from the relative positional relationship between the positioning mark frame in the template image and the positioning mark of the given image ...

Embodiment 2

[0060] The number of the anchors of the given image and the number of the anchor frames of the template image are preset, and the anchor frames correspond to the anchors one by one.

[0061] Determine the first relative positional relationship between the positioning mark frame in the template image and the positioning mark in the given image, and determine and segment the object to be recognized in the given image according to the first relative positional relationship and the position of the slice frame in the template image The corresponding area may specifically include: calculating the second relative positional relationship between each positioning mark frame in the template image and the corresponding positioning mark in the given image; using each second relative positional relationship to calculate all positioning positions of the given image According to the optimal relative positional relationship and the position of the slice frame in the template image, determine a...

Embodiment 3

[0067] Determine the first relative positional relationship between the positioning mark frame in the template image and the positioning mark in the given image, and determine and segment the object to be recognized in the given image according to the first relative positional relationship and the position of the slice frame in the template image The corresponding area includes: determining the third relative positional relationship between the reference point of the positioning mark frame in the template image and the reference point of the positioning mark in the given image; according to the third relative positional relationship, the given image and the template image The zoom ratio and the position of the slice frame in the template image determine and segment the area corresponding to the object to be recognized in the given image. Specifically, the third relative positional relationship may be the coordinate offset between the reference point of the anchor frame in the t...

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 object recognition method and device, and relates to the technical field of image recognition. A specific embodiment of the method comprises the following steps: determining the position of a localizer in a given image, wherein the localizer is used for locating an object to be identified in the given image; according to the position of the localizer, segmenting an area corresponding to the object to be identified from the given image to obtain a slice image; and inputting the slice image into a classifier to obtain an identification result of the to-be-identified object. According to the embodiment, the object can be accurately recognized under the condition that the content of the object is diversified, fuzzy matching is not needed, and recognition steps are simplified.

Description

technical field [0001] The present invention relates to the technical field of image recognition, in particular to an object recognition method and device. Background technique [0002] Due to the diversity of the content of the object, it will lead to various results for the same object. Taking the identification of administrative divisions as an example, first of all, when people fill in the administrative divisions, some of the smallest units are written as counties, and some of the smallest units are written as villages. Secondly, the national administrative divisions also often change, for example, "Xuanwu District, Beijing" becomes "Xicheng District, Beijing". In the identification of the administrative division column, there are the following particularities. First, the user fills in irregularly. Some write districts, some write counties, and some omit provinces and directly write cities, etc. Second, the national administrative divisions often Change, resulting in t...

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): G06K9/62G06T7/00G06T7/11G06T7/70G06N3/04G06N3/08
CPCG06T7/0002G06T7/11G06T7/70G06N3/08G06N3/047G06N3/045G06F18/2415
Inventor 李虎冯程郑邦东
Owner CHINA CONSTRUCTION BANK
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