Convolutional neural network-based visible light image weather recognition method

A convolutional neural network and weather recognition technology, applied in the field of visible light image weather recognition, can solve the problems of weather information discrimination defects, interfere with the acquisition of deep abstract information, etc., to improve the classification and recognition effect, improve recognition accuracy, and increase perception opportunities. Effect

Inactive Publication Date: 2018-11-23
SHANGHAI JIAO TONG UNIV
View PDF7 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the existing CNN model is more effective in discriminating information such as the more vivid foreground and background in the image, but this information will interfere with the acquisition of deep abstract information, so there are certain defects in discriminating abstract weather information

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
  • Convolutional neural network-based visible light image weather recognition method
  • Convolutional neural network-based visible light image weather recognition method
  • Convolutional neural network-based visible light image weather recognition method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] Below in conjunction with accompanying drawing and experimental sample the present invention is described in detail:

[0031] The first step is to obtain the original data of the visible light weather image, including the image data and the weather parameters of the target image;

[0032] The second step is to perform preprocessing operations on the data, including image reconstruction and image enhancement operations on images with poor imaging conditions;

[0033] The third step is to use the Mask R-CNN method to extract the region of interest from the visible light weather image. Through the pre-trained Mask R-CNN model, extract the largest foreground in visible light weather images, such as vehicles, buildings, etc. Cut off the edge of the foreground and extract it as the region of interest. Finally, the image is segmented into a foreground part and a background part without foreground.

[0034] Among them, Mask R-CNN is a convolutional neural network image recog...

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 convolutional neural network-based visible light image weather recognition method. According to the method, a convolutional neural network and visible light image data are combined; and through a method of constructing a training set after extracting a region of interest, the sensitivity, to image superficial layer information, of the network is reduced, the judgement toabstract weather information is enhanced, the training precision is improved, so that a more correct visible light weather recognition model is obtained and the recognition precision is improved.

Description

technical field [0001] The invention relates to the technical field of image processing and pattern recognition, in particular to a weather recognition method for visible light images based on convolutional neural networks. Background technique [0002] In modern road traffic, severe weather conditions will have a great impact on urban traffic. Fog, snow, and other extreme weather conditions will cause a significant drop in visibility and slippery road surfaces, which may not only cause traffic jams and reduce transportation speed. If it is lowered, it may also cause accidents such as car accidents. At this time, through the real-time monitoring of the weather and environment, and the comprehensive use of road conditions and traffic flow information, the use of traffic lights can effectively reduce the occurrence of adverse traffic conditions and improve travel efficiency under severe weather conditions. However, the workload of manually judging the weather conditions at ea...

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): G06K9/00G06K9/46G06N3/04
CPCG06V20/13G06V10/462G06N3/045
Inventor 李元祥施雨舟李子佳陈辰
Owner SHANGHAI JIAO TONG UNIV
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