Accurate positioning and recognition method for train number on basis of deep learning

A technology of precise positioning and deep learning, applied in the field of intelligent transportation, can solve the problems such as the inability to locate the vehicle number, achieve real-time accurate positioning and accurate identification, accurate sample positioning, and avoid the effect of cumulative error

Pending Publication Date: 2020-04-10
BEIJING JIAOTONG UNIV +1
View PDF2 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This kind of method is easy to locate the vehicle number into two separate parts, and cannot realize the positioning of the complete vehicle number

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
  • Accurate positioning and recognition method for train number on basis of deep learning
  • Accurate positioning and recognition method for train number on basis of deep learning
  • Accurate positioning and recognition method for train number on basis of deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0037] figure 1 A schematic flow chart of a method for precise positioning and identification of train numbers based on deep learning provided in this embodiment, refer to figure 1 , the method includes:

[0038] S1 collects the panorama of the train, and adjusts the size of the panorama.

[0039] Collect the panorama of the train captured by the surveillance camera, and scale the panorama to a picture with a height of 600 pixels or a picture with a width of 1000 pixels in the same proportion.

[0040] The specific method is: input a panorama of any size, scale the width and height of the input panorama in the same proportion until the height is 600 pixels, if the width exceeds 1000 pixels at this time, continue to shrink the input panorama until the panorama’s The width is 1000 pixels.

[0041] S2 Constructing a train number location network, and using the adjusted panorama as a training set to train the train number location network.

[0042] The train number location ne...

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 provides an accurate positioning and recognition method for a train number on the basis of deep learning. The accurate positioning and recognition method comprises the steps of: collecting a train panoramic image, and carrying out size adjustment on the train panoramic image; constructing a train number positioning network, and taking an adjusted panoramic image as a training set totrain the train number positioning network; training a train number identification network according to the train number area output by the train number positioning network; performing size adjustmenton a to-be-identified train panoramic image, and inputting an adjusted train panoramic image into a trained train number positioning network so as to obtain an accurately-positioned train number area; and inputting the train number area into a trained train number identification network for identification so as to obtain a train number digital identification result. With the accurate positioningand recognition method provided by the invention, the train number sequence with any length can be processed; the defects of low positioning accuracy and difficulty in distinguishing small-size vehicle numbers for manual features in a complex scene and a conventional deep learning method are overcome; meanwhile, character segmentation is not involved; and overall recognition is achieved.

Description

technical field [0001] The invention relates to the technical field of intelligent transportation, in particular to a method for accurately locating and identifying train numbers based on deep learning. Background technique [0002] The automation of train inspection and cargo inspection operations is an important goal of railway informatization. As one of the basic tasks of inspection automation, the positioning and identification of train numbers can automatically record the number information and avoid consuming huge manpower and material resources. When an abnormal condition of a freight train or a passenger train is detected, the train number information is automatically used as a sign of the identity of the train, linked with the safety status information of the equipment, and an alarm is sent to the control center. Therefore, train number identification is used in train inspection and It is particularly important in the automation of cargo inspection operations. [0...

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): G06K9/32G06K9/62
CPCG06V20/62G06V30/10G06F18/241G06F18/2415
Inventor 邹琪艾鑫罗常津杨文冠丁正刚胡宸瀚周通阳勇杰徐嫣
Owner BEIJING JIAOTONG 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