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

Road surface abandoned object detection method based on deep convolutional network

A technology of deep convolution and detection method, which is applied in the field of detection of road residues based on deep convolutional neural network, can solve the problems of inability to effectively manage existing video resources, and achieves a low-cost, low-cost hardware investment. Effect

Active Publication Date: 2017-06-13
NANJING UNIV
View PDF5 Cites 34 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The problem to be solved by the present invention is: as the number of road monitoring videos increases sharply, it is impossible to achieve effective management of existing video resources only by manual work

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
  • Road surface abandoned object detection method based on deep convolutional network
  • Road surface abandoned object detection method based on deep convolutional network
  • Road surface abandoned object detection method based on deep convolutional network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] The invention introduces deep learning into road event recognition and improves it, which can significantly improve the accuracy of road event recognition. Considering that the background modeling method is not suitable for static object detection, and the pavement remnants are difficult to use the prior model to construct the training set. The invention establishes the detection of road remnants based on deep convolutional neural network Deep-CNN. The detection point of the mobile terminal is a road camera, and the detection point of the mobile terminal obtains image information through the camera, and the CNN neural network is used to analyze the obtained image. The present invention Divide the pavement ROI area into multiple networks, build a pavement-non-pavement recognition model, and reversely identify static targets such as highway pavement remnants and road spills through non-pavement grids.

[0029] The present invention builds the layered road event recognitio...

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 road surface abandoned object detection method based on a deep convolutional network. Mobile terminal detection points serve as road cameras and acquire image information through cameras, and deep learning is introduced into road surface event recognition and is improved so as to significantly improve road event recognition accuracy. According to the method, acquired images are analyzed through the convolutional neural network, and target detection with mobile cameras and static images is achieved; a road surface ROI is divided into multiple meshes, a road surface-non-road surface recognition model is built, and static targets such as highway surface abandoned objects and thrown objects are reversely recognized through non-road surface meshes. The method is applied to non-real-time tasks like detection of road surface abandoned objects and thrown objects, the characteristics and advantages of the mobile internet are fully utilized, and detection of road surface events like road surface abandoned objects with a high region coverage rate is achieved at a low cost.

Description

technical field [0001] The invention belongs to the technical field of video detection of depth features, relates to the detection of road events, detects the remnants of the road based on the Deep-CNN road reverse recognition model, and performs data analysis and data mining on the detection results. A neural network (Deep-CNN) detection method for pavement remnants. Background technique [0002] Road accidents, traffic congestion, and environmental pollution are common problems faced by the development of today's road traffic. Road traffic safety is worrisome. Road informatization and Intelligent Transportation System (ITS) are effective means to improve the utilization efficiency of road facilities, alleviate traffic congestion, and reduce the incidence of traffic accidents. Use computer vision technology to sense traffic flow parameters such as road speed and flow in real time, provide real-time road conditions, and combine historical data to predict road network traffi...

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): G06K9/00G06K9/32G06K9/34G06K9/62G06N3/04
CPCG06V20/41G06V10/25G06V10/267G06N3/045G06F18/214
Inventor 阮雅端高妍张宇杭张园笛陈启美
Owner NANJING UNIV
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