Method and system for identifying train number and train type, and method and system for security inspection

Patent metadata

Field	Value
Publication number	US-10204268-B2
Application number	US-201615280113-A
Country	US
Kind code	B2
Filing date	Sep 29, 2016
Priority date	Dec 29, 2015
Publication date	Feb 12, 2019
Grant date	Feb 12, 2019

Abstract

The present disclosure provides a method and system for identifying a train number and train type. The method includes: continuously photographing a train under inspection by using a linear-array camera in motion relative to the train under inspection, and generating a plurality of partial images of the train; splicing the plurality of partial images of the train; correcting distortion of the spliced image; identifying a train number from the corrected image; wherein the correcting distortion of the spliced image includes: extracting a contour of a wheel from the spliced image; obtaining a ratio between a horizontal diameter and a vertical diameter of the wheel from the contour; if the ratio is greater than a first preset threshold, horizontally compressing the spliced image according to the ratio; and if the ratio is smaller than a second preset threshold, horizontally stretching the spliced image.

First claim

What is claimed is:

1. A train number identification system, comprising:
a processor; and
a memory storing instructions executable by the processor,
wherein the processor is configured to:
obtain a plurality of partial images of a train under inspection which are generated by continuously photographing the train wider inspection using a linear-array camera in motion relative to the train under inspection;
splice the plurality of partial images of the train to obtain a spliced image;
correct distortion of the spliced image to obtain a corrected image; and
identify a train number from the corrected image;
wherein the correcting of distortion of the spliced image comprises:
extracting a contour of a wheel from the spliced image;
obtaining a ratio between a horizontal diameter and a vertical diameter of the wheel from the contour,
if the ratio is greater than a first preset threshold, horizontally compressing the spliced image according to the ratio; and if the ratio is smaller than a second preset threshold, horizontally stretching the spliced image.

2. The train number identification system according to claim 1 , wherein the processor is configured to obtain the plurality of partial images photographed by the linear-array camera at a preset photographing frequency.

3. The train number identification system according to claim 1 , wherein the processor is configured to:
cut a train number region from the corrected image to obtain an image of the train number region;
perform smoothing and binary processing on the image of the train number region;
identify the train number using a character identifying engine; and
automatically correct the identified train number according to a rule for defining train numbers.

4. A train type identification system, comprising:
a train number identification module configured to identify a train number of a train; and
a train type determination module configured to determine a type of the train according to the train number,
wherein the train number identification module is a train number identification system;
wherein the train number identification system comprises:
a processor; and
a memory storing instructions executable by the processor;
wherein the processor is configured to:
obtain a plurality of partial images of a train under inspection which are generated by continuously photographing the train under inspection using a linear-array camera in motion relative to the train under inspection;
splice the plurality of partial images of the train to obtain a spiked image;
correct distortion of the spliced image to obtain a corrected image; and
identify a train number from the corrected image;
wherein the correcting of distortion of the spliced image comprises:
extracting a contour of a wheel from the spliced image;
obtaining a ratio between a horizontal diameter and a vertical diameter of the wheel from the contour;

if the ratio is greater than a first preset threshold, horizontally compressing the spliced image according to the ratio; and if the ratio is smaller than a second preset threshold, horizontally stretching the spliced image.

5. The train type identification system according to claim 4 , wherein the train type determination module is configured to search a type from a data base or a data list using the train number.

6. The train type identification system according to claim 4 , wherein the train type determination module is configured to directly determine the type of the train from the train number according to a rule for defining train numbers.

7. The train type identification system according to claim 4 , wherein the type of the train is any one of a locomotive train, a freight train, and a passenger train.

8. The train type identification system according to claim 4 , further comprising:
a coupler identification module configured to identify positions of couplers of the train; and
an image partitioning module configured to partition an image of the train into a plurality of sub-images of the train according to the positions of the couplers.

9. A system for performing security inspection on a train, comprising:
a train type identification system; and
a radiation control module configured to, according to a type of a train under inspection as determined by the train type identification system, control a ray source to irradiate the train under inspection with a first dosage or not irradiate the train under inspection when the train under inspection is a locomotive train or a passenger train, and control the ray source to irradiate the train under inspection with a second dosage when the train under inspection is a freight train, wherein the first dosage is smaller than the second dosage;
wherein the train type identification system comprises:
a train number identification module configured to identify a train number of a train; and
a train type determination module configured to determine a type of the train according to the train number,
wherein the train number identification module is a train number identification system;
wherein the train number identification system comprises:
a processor; and
a memory storing instructions executable by the processor;
wherein the processor is configured to:
obtain a plurality of partial images of a train under inspection which are generated by continuously photographing the train under inspection using a linear-array camera in motion relative to the train under inspection;
splice the plurality of partial images of the train to obtain a spliced image;
correct distortion of the spliced image to obtain a corrected image; and
identify a train number from the corrected image;
wherein the correcting of distortion of the spliced image comprises:
extracting a contour of a wheel from the spliced image;
obtaining a ratio between a horizontal diameter and a vertical diameter of the wheel from the contour;

if the ratio is greater than a first preset threshold, horizontally compressing the spliced image according to the ratio; and if the ratio is smaller than a second preset threshold, horizontally stretching the spliced image.

10. The system according to claim 9 , wherein the train type determination module is configured to search a type from a data base or a data list using the train number.

11. The system according to claim 9 , wherein the train type determination module is configured to directly determine the type of the train from the train number according to a rule for defining train numbers.

12. The system according to claim 9 , wherein the type of the train is any one of a locomotive train, a freight train, and a passenger train.

13. The system according to claim 9 , further comprising:
a coupler identification module configured to identify positions of couplers of the train; and
an image partitioning module configured, to partition an image of the train into a plurality of sub-images of the train according to the positions of the couplers.

Assignees

• Nuctech Co Ltd

Inventors

• Xu Yanwei
• Yu Weifeng

Classifications

• G06V30/168 Primary

  Smoothing or thinning of the pattern; Skeletonisation · CPC title

• G06F18/00 Primary

  Pattern recognition · CPC title

• G06V30/10

  Character recognition · CPC title

• H04N2209/045

  using mosaic colour filter · CPC title

• G06K9/00624 Primary

  Physics · mapped topic