Anti-complex scene interference manhole cover category automatic detection method and system

An automatic detection and complex scene technology, applied in character and pattern recognition, instruments, computer components, etc., can solve problems such as error-prone, low efficiency, high risk, etc., to reduce missed detection rate, high-efficiency detection effect, improve The effect of detection accuracy

Active Publication Date: 2020-11-10
SOUTH CHINA UNIV OF TECH
View PDF2 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the inefficiency, error-proneness and high risk of the existing manhole cover category detection method, the present invention provides an automatic manhole cover category detection method that is resistant to complex scene interference

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
  • Anti-complex scene interference manhole cover category automatic detection method and system
  • Anti-complex scene interference manhole cover category automatic detection method and system
  • Anti-complex scene interference manhole cover category automatic detection method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0070] An automatic detection method for manhole cover types against complex scene interference, comprising the following steps:

[0071] S1. Image collection and classification: A large number of pictures containing different types of manhole covers are intercepted from surveillance videos of real scenes to generate multi-category training data sets and test data sets.

[0072] Specifically, the image collection and classification steps include:

[0073] S11. Collect different types of manhole covers in the monitoring scene, and initially divide them into three categories: complete, damaged and lost;

[0074] S12. Download related pictures of different types of manhole covers through crawlers, wherein the crawler program can automatically remove abnormal pictures;

[0075] S13. Perform secondary screening on the downloaded manhole cover-related pictures, and remove pictures with low resolution that are difficult for human eyes to obtain a manhole cover image set;

[0076] S...

Embodiment 2

[0113] An automatic detection system for manhole covers that is resistant to complex scene interference, including:

[0114] Image collection and classification module: used to intercept a large number of pictures containing different manhole cover categories from the surveillance video of the real scene, and generate multi-category manhole cover training data sets and test data sets;

[0115] Image labeling module: used to specify a class name for each manhole cover category, and calibrate all manhole cover image sets to obtain an xml file that records the manhole cover category and its location information in the manhole cover picture;

[0116] The data expansion module is used to perform data expansion for different directions, positions, zoom ratios, and brightness of the training data set, and perform random division according to the best ratio;

[0117] Manhole cover detection model module: ResNet-101 model is used as the network skeleton, and a series of convolutional l...

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 belongs to the technical field of deep learning and computer smart city monitoring, and discloses an anti-complex scene interference manhole cover category automatic detection method, which comprises the steps of S1, performing image collection and classification; S2, carrying out image annotation; S3, carrying out data expansion; S4, adopting a ResNet-101 model as a network skeleton, adding a series of convolution layers, pooling layers and full connection layers from large to small on the ResNet-101 model for predicting category information and position information of a manholecover in a picture, and constructing an R-FCN manhole cover detection model; S5, obtaining an optimal R-FCN manhole cover detection model through iterative training; and S6, identifying the picture or the video to obtain the category and position information of the manhole cover in the picture or the video. According to the invention, the damage of the manhole cover can be detected more efficiently by using the method based on deep learning. The invention further provides an anti-complex scene interference manhole cover category automatic detection system.

Description

technical field [0001] The invention relates to the technical field of deep learning and computer smart city monitoring, in particular to an automatic detection method and system for manhole cover types that are resistant to complex scene interference. Background technique [0002] With the development of the urban traffic system and the continuous increase in the number of vehicles, people pay more and more attention to road traffic safety. Manhole covers exist in the streets and alleys of the city. Timely monitoring of the loss and damage of manhole covers will help to avoid traffic accidents. accidents and reduce hazards such as injury and death. In the investigation method of road surface manhole cover damage, traditional manual detection method is adopted, that is, each manhole cover on the urban road surface is checked manually, and the existence of damaged manhole cover is recorded and found. This method not only requires a lot of manpower and material resources, but...

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): G06K9/00G06K9/32G06K9/46G06K9/62
CPCG06V20/41G06V20/52G06V10/25G06V10/40G06F18/2415G06F18/214Y02T10/40
Inventor 黄翰钟胜杰杨忠明许红涛邹梦良郝志峰
Owner SOUTH CHINA UNIV OF TECH
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