CN107284479B - Locomotive number identification method and device based on image processing - Google Patents
Locomotive number identification method and device based on image processing Download PDFInfo
- Publication number
- CN107284479B CN107284479B CN201710592611.3A CN201710592611A CN107284479B CN 107284479 B CN107284479 B CN 107284479B CN 201710592611 A CN201710592611 A CN 201710592611A CN 107284479 B CN107284479 B CN 107284479B
- Authority
- CN
- China
- Prior art keywords
- image
- locomotive
- video stream
- frame
- number identification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000003137 locomotive effect Effects 0.000 title claims abstract description 100
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000010586 diagram Methods 0.000 description 5
- 238000013481 data capture Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L25/00—Recording or indicating positions or identities of vehicles or trains or setting of track apparatus
- B61L25/02—Indicating or recording positions or identities of vehicles or trains
- B61L25/04—Indicating or recording train identities
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Image Analysis (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The invention discloses a locomotive number identification method and device based on image processing, which are used for solving the problems of low locomotive number identification speed and low identification degree. The principle is as follows: by processing the background image, the interference data in the acquired image data is quickly eliminated until the complete train number of the locomotive is identified. In the locomotive number identification process, effective images are only identified for the video stream corresponding to each locomotive, so that the identification speed and the identification degree are higher than those of the prior art.
Description
Technical Field
The invention relates to the technical field of locomotive number identification, in particular to a locomotive number identification method and device based on image processing.
Background
In the current railway management, the car passing management is carried out in an automatic car number identification system of a professional locomotive engineer, and the working principle is as follows: the locomotive passes through the assigned position, triggers wheel sensor, and the signal receiving integrated circuit board that crosses the car receives the incoming telegram signal and transmits to the host computer of locomotive service station. The host computer sends a control instruction to the camera and the light, and the camera starts to take pictures. Software comprising a car number recognition algorithm in the host machine intelligently recognizes the collected pictures and counts recognition results, and a corresponding car number is obtained according to the counting results.
The host of the locomotive station executes a car number identification algorithm on the acquired video stream, but not all images in the acquired video stream are valid car number images, and the invalid video images account for the vast majority of the whole video stream, so that the time for identifying the car number is long, and the efficiency is not high.
disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, the present invention is directed to a locomotive car number identification method and apparatus based on image processing, which is used to solve the problem of low identification speed of the car number identification device in the car number identification process, and other related technical problems.
In order to achieve the above objects and other related objects, the present invention provides the following technical solutions:
According to a first aspect of the invention:
A locomotive number identification method based on image processing comprises the following steps: searching a frame of image containing a background in the content close to the head side/tail side of each locomotive in the vehicle-passing video stream; starting a train number identification program, and starting train number identification from the position of the frame image in the train passing video stream to the tail side/head side of the locomotive in the train passing video stream until the complete train number of the locomotive is identified.
Further, the method for searching for a frame of image containing background in the content close to the head side/tail side of each locomotive in the vehicle video stream comprises the following steps: the image is looked up according to the dichotomy.
Further, the step of finding the image according to a dichotomy comprises: selecting a start frame image or a last frame image of a passing video stream of each locomotive as an anchor image; comparing an image obtained by bisection search of the passing video stream of the locomotive with the anchor image, and stopping searching when the image and the anchor image meet preset conditions, wherein the image is a frame of image which is close to the head side/tail side of each locomotive and contains a background.
further, when the car number identification program is started to identify the car number, the identification mode specifically includes: and identifying the images in the passing video stream frame by frame.
Further, when the car number identification program is started to identify the car number, the identification mode specifically includes: and sequentially skipping and identifying the images in the vehicle passing video stream, and identifying the images in the vehicle passing video stream frame by frame when the images contain characters.
According to another aspect of the invention
Still provide a locomotive car number recognition device based on image processing, include: the image processing module is used for searching a frame of image which is close to the content of the head side/tail side of each locomotive in the passing video stream and contains the background; and the car number identification module starts a car number identification program, starts car number identification from the position of the frame image in the passing video stream to the tail side/head side of the locomotive in the passing video stream until the complete car number of the locomotive is identified.
Further, the image processing module is specifically configured to: and searching a frame image of the contents close to the head side/tail side of each locomotive in the passing video stream, wherein the frame image contains the background according to a dichotomy.
Further, the image processing module specifically includes: the anchor image determining unit is used for selecting a start frame image or a last frame image of a passing video stream where each locomotive is located as an anchor image; and the image searching unit is used for comparing an image obtained by bisection searching of the passing video stream of the locomotive with the anchor image, stopping searching when the image and the anchor image meet preset conditions, wherein the image is a frame of image which is close to the head side/tail side of each locomotive and contains a background.
further, when the car number identification module is used for enabling the car number identification program to identify the car number, the identification mode specifically includes: and identifying the images in the passing video stream frame by frame.
Further, when the car number identification module is used for enabling the car number identification program to identify the car number, the identification mode specifically includes: and sequentially skipping and identifying the images in the vehicle passing video stream, and identifying the images in the vehicle passing video stream frame by frame when the images contain characters.
the invention has the following beneficial effects: the method comprises the steps of finding a frame of image, which is close to a head side or a tail side and contains a background, in a video stream through image processing, quickly eliminating interference data in collected image data by taking the frame of image as a reference, then starting a car number identification program to identify a car number, and only identifying effective images of the video stream corresponding to each locomotive in the car number identification process of the locomotive, so that the identification speed and the identification degree are higher than those of the prior art.
Drawings
Fig. 1 is a flowchart illustrating a locomotive number identification method based on image processing according to an embodiment of the present invention.
Fig. 2 is a schematic diagram showing the image composition of a video stream of passing vehicles according to the present invention.
Fig. 3 is a flow chart of another embodiment of the method for identifying a locomotive number based on image processing according to the present invention.
Fig. 4 is a block diagram showing a first configuration of the locomotive number identification device based on image processing according to the present invention.
Fig. 5 is a block diagram showing a structure of a locomotive number recognition device based on image processing according to the present invention.
description of the reference numerals
400 device
410 image processing module
411 image search unit
412 image recognition unit
420 image recognition module
S101 to S102
S201 to S202
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
Referring to fig. 1, the present embodiment provides a flowchart of a method for identifying a locomotive number based on image processing, and as shown in the figure, the method for identifying a locomotive number specifically includes the following steps:
S101, searching a frame of image containing a background in the content close to the head side/tail side of each locomotive in the passing video stream;
s102, starting a train number identification program, and starting train number identification from the position of the frame image in the train passing video stream to the tail side/head side of the locomotive in the train passing video stream until the complete train number of the locomotive is identified.
The method starts to identify the car number of the background image close to the head side or the tail side by starting a car number identification program, so that the complete car number can be identified only by identifying partial images (namely effective images) in the video stream corresponding to the locomotive.
the train-passing video stream, also referred to as a video stream for short, refers to video data obtained by acquiring video images of a passing train through a video acquisition device when the train passes through. Specifically, the video data includes a plurality of frames of images arranged in time sequence, and in the plurality of frames of images, the content of partial images is a locomotive body, the content of partial images is a background, and the content of partial images is a combination of the locomotive body and the background.
fig. 2 is a schematic diagram of the image composition of a video stream of passing vehicles according to the present invention, and as shown, the abscissa represents the time t (time); the ordinate represents the frame image f (frame), and it is clear that a video stream is temporally arranged by images of many frames. For ease of illustration, an image composition analysis of the video stream of only two locomotives is shown in FIG. 2, with T1 representing video stream data acquired for a first locomotive and T2 representing video stream data acquired for a second locomotive.
Specifically, the video stream is divided based on the end of video acquisition of one locomotive and the start of video acquisition of the next locomotive. For example, a video stream may include: captured video during the time period between when one locomotive begins to pass the video stream data capture device and when the next locomotive arrives at the video stream data capture device.
More specifically, if the video data acquisition is started when the locomotive at the head of the train just arrives at or passes through the video data acquisition device, the content of the multi-frame image of the video stream acquired at the time t1 only contains the locomotive body; if the video data acquisition is started when the locomotive at the head of the train is close to the video data acquisition equipment, the content of the multi-frame image of the acquired video stream at the time t1 should also contain a background image. It is clear that the video stream acquired in fig. 2 belongs to the former type of video stream.
Specifically, in the video stream acquired at time T1, according to the content included in each frame of image, the method may further include: the image content acquired at time t1 includes a plurality of frame images of the locomotive body and the image content acquired at time t2 includes a plurality of frame images of the background.
Of course, in the multi-frame image acquired in the time t1, the following steps can be further divided: the contents include the multi-frame images t1a and t1b of the locomotive number characters, and the multi-frame image t1c of the contents not including the locomotive number characters.
Based on the above analysis of the video stream, it can be seen that only the multi-frame images t1a and t1b, which are acquired during the time t1 and whose contents include the locomotive car number characters, are necessary for car number identification. The reason that the conventional car number identification efficiency is not high is that images except t1a and t1b are also identified one by one.
The principle of the invention is as follows: firstly, finding a background image close to the tail side of each locomotive, namely an image with one frame of content as a background in video stream data acquired by video data acquisition equipment during the period from the time when all locomotives pass through the video data acquisition equipment to the time when the next locomotive reaches the video data acquisition equipment; since the background image is closest to the multi-frame image t1b containing the locomotive car number characters, the background image is used as the starting point of the car number identification, and the image is identified towards the side of the multi-frame image t1b containing the locomotive car number characters until the complete car number is identified. Therefore, the complete train number of the locomotive can be identified only by identifying part of images, and the identification efficiency is far higher than that of the existing train number identification method.
Obviously, it is also possible to find the background image near the locomotive side of each locomotive, and then recognize the image toward the side of the multi-frame image t1a whose content contains the locomotive number characters until the complete car number is recognized.
In a preferred embodiment, the method for searching for a frame of image containing background in the video stream of the passing vehicle near the head side/tail side of each locomotive comprises: the image is looked up according to the dichotomy.
In a preferred embodiment, referring to fig. 3, the step of finding the image according to a dichotomy comprises:
S201, selecting a start frame image or a last frame image of a passing video stream where each locomotive is located as an anchor image;
S202, comparing an image obtained by bisection searching of the passing video stream of the locomotive with the anchor image, and stopping searching when the image and the anchor image meet preset conditions, wherein the image is a frame of image which is close to the head side/tail side of each locomotive and contains a background.
wherein, the anchor image refers to a standard image for comparison.
After finding a frame of image with the content including a background, starting a train number identification program to identify the train number, specifically, if the frame of image is positioned at the head side of a locomotive in a video stream, performing the train number identification by taking the frame of image as a starting point to an image in the direction of the tail side; similarly, if the frame image is located at the tail side of the locomotive in the video stream, the car number identification is performed by taking the frame image as an image of the starting point in the direction of the head side.
In a preferred scheme, the method for identifying the vehicle number of the passing video stream by using the vehicle number identification program specifically comprises the following steps: and identifying the images in the passing video stream frame by frame. Therefore, from the frame of image to the identification of the complete car number, the car number identification program only needs to identify a small number of images, so that the identification speed is higher in the prior art.
in another preferred scheme, the method for identifying the vehicle number of the passing video stream by using the vehicle number identification program specifically comprises the following steps: and sequentially skipping and identifying the images in the vehicle passing video stream, and identifying the images in the vehicle passing video stream frame by frame when the images contain characters. Since more invalid images are still included between the frame of image with the searched content including the background and the identification of the complete car number corresponding image, for example, some images with the content only being the car body exist, the number of images to be identified by the car number identification program can be further reduced through the frame skipping identification.
referring to fig. 4, the present embodiment further provides a block diagram of a first apparatus for implementing the method, as shown in the figure, the apparatus 400 includes an image processing module 410 and a car number identification module 420, the image processing module 410 is configured to find a frame of image in the passing video stream, where the content near the head side/tail side of each locomotive includes a background; and the car number identification module 420 starts a car number identification program, starts car number identification from the position of the frame image in the passing video stream to the tail side/head side of the locomotive in the passing video stream until the complete car number of the locomotive is identified.
Preferably, the image processing module 410 is specifically configured to: and searching a frame image of the contents close to the head side/tail side of each locomotive in the passing video stream, wherein the frame image contains the background according to a dichotomy.
Preferably, referring to fig. 5, the image processing module 410 specifically includes: an anchor image determining unit 411, configured to select a start frame image or a last frame image of a passing video stream where each locomotive is located, as an anchor image; the image searching unit 412 compares an image obtained by bisection searching of the passing video stream of the locomotive with the anchor image, and stops searching when the image and the anchor image meet preset conditions, wherein the image is a frame of image with a background included in the content close to the head side/tail side of each locomotive.
Preferably, when the car number identification module is used for enabling the car number identification program to identify the car number, the identification mode specifically includes: and identifying the images in the passing video stream frame by frame.
preferably, when the car number identification module is used for enabling the car number identification program to identify the car number, the identification mode specifically includes: and sequentially skipping and identifying the images in the vehicle passing video stream, and identifying the images in the vehicle passing video stream frame by frame when the images contain characters.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (6)
1. A locomotive number identification method based on image processing is characterized by comprising the following steps:
Searching a frame of image containing a background in the content close to the head side/tail side of each locomotive in the vehicle-passing video stream;
Starting a train number identification program, and starting train number identification from the position of the frame image in the train passing video stream to the tail side/head side of the locomotive in the train passing video stream until the complete train number of the locomotive is identified;
the method for searching the frame image of the content containing the background close to the head side/tail side of each locomotive in the vehicle-passing video stream comprises the following steps: searching the image according to a dichotomy;
The step of finding the image according to a dichotomy comprises:
Selecting a start frame image or a last frame image of a passing video stream of each locomotive as an anchor image;
Comparing an image obtained by bisection search of the passing video stream of the locomotive with the anchor image, and stopping searching when the image and the anchor image meet preset conditions, wherein the image is a frame of image which is close to the head side/tail side of each locomotive and contains a background.
2. the image processing-based locomotive number identification method according to claim 1, wherein: when the car number identification program is started to identify the car number, the identification mode specifically comprises the following steps: and identifying the images in the passing video stream frame by frame.
3. The image processing-based locomotive number identification method according to claim 1, wherein: when the car number identification program is started to identify the car number, the identification mode specifically comprises the following steps:
and sequentially skipping and identifying the images in the vehicle passing video stream, and identifying the images in the vehicle passing video stream frame by frame when the images contain characters.
4. A locomotive car number recognition device based on image processing, comprising:
The image processing module is used for searching a frame of image which is close to the content of the head side/tail side of each locomotive in the passing video stream and contains the background;
the train number identification module starts a train number identification program, starts train number identification from the position of the frame image in the train passing video stream to the tail side/head side of the locomotive in the train passing video stream until the complete train number of the locomotive is identified;
The image processing module is specifically configured to: searching a frame of image containing a background in the content close to the head side/tail side of each locomotive in the passing video stream according to a bisection method;
The image processing module specifically comprises:
The anchor image determining unit is used for selecting a start frame image or a last frame image of a passing video stream where each locomotive is located as an anchor image;
And the image searching unit is used for comparing an image obtained by bisection searching of the passing video stream of the locomotive with the anchor image, stopping searching when the image and the anchor image meet preset conditions, wherein the image is a frame of image which is close to the head side/tail side of each locomotive and contains a background.
5. The image processing-based locomotive number identification device according to claim 4, wherein: when the car number identification module is used for starting the car number identification program to identify the car number, the identification mode specifically comprises the following steps: and identifying the images in the passing video stream frame by frame.
6. the image processing-based locomotive number identification device according to claim 4, wherein: when the car number identification module is used for starting the car number identification program to identify the car number, the identification mode specifically comprises the following steps:
And sequentially skipping and identifying the images in the vehicle passing video stream, and identifying the images in the vehicle passing video stream frame by frame when the images contain characters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710592611.3A CN107284479B (en) | 2017-07-19 | 2017-07-19 | Locomotive number identification method and device based on image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710592611.3A CN107284479B (en) | 2017-07-19 | 2017-07-19 | Locomotive number identification method and device based on image processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107284479A CN107284479A (en) | 2017-10-24 |
| CN107284479B true CN107284479B (en) | 2019-12-17 |
Family
ID=60102275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710592611.3A Active CN107284479B (en) | 2017-07-19 | 2017-07-19 | Locomotive number identification method and device based on image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107284479B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108108710A (en) * | 2017-12-29 | 2018-06-01 | 重庆微标科技股份有限公司 | By obtaining effective vehicle image of crossing come the quick license number identified on vehicle device vehicle |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201102555Y (en) * | 2007-08-06 | 2008-08-20 | 深圳市远望谷信息技术股份有限公司 | Wagon number recognition system for railway |
| CN101453575B (en) * | 2007-12-05 | 2010-07-21 | 中国科学院计算技术研究所 | Video subtitle information extracting method |
| CN202098436U (en) * | 2011-05-31 | 2012-01-04 | 合肥安迅铁道应用技术有限公司 | Motor train unit identification system based on video wagon number identification |
| WO2015141586A1 (en) * | 2014-03-20 | 2015-09-24 | 株式会社日立国際電気 | Intra-formation network system, intra-formation network construction method, and management device |
| CN105667545B (en) * | 2015-12-31 | 2017-12-29 | 重庆微标科技股份有限公司 | A kind of railway locomotive Train number recognition method and system |
-
2017
- 2017-07-19 CN CN201710592611.3A patent/CN107284479B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107284479A (en) | 2017-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108229386B (en) | Method, apparatus, and medium for detecting lane line | |
| CN110473406B (en) | Search support system, search support device, and search support method | |
| CN102110369B (en) | Jaywalking snapshot method and device | |
| US20220157172A1 (en) | System, program, and method for detecting information on a person from a video of an on-vehicle camera | |
| CN110648539B (en) | In-vehicle device and control method | |
| CN111242943B (en) | Image processing method, image processing apparatus, storage medium, and image processing device | |
| CN109509274B (en) | Method and device for reducing environmental information data of a driver assistance system | |
| CN107284479B (en) | Locomotive number identification method and device based on image processing | |
| CN107393308A (en) | A kind of method, apparatus and managing system of car parking for identifying car plate | |
| CN113895481A (en) | Train positioning and tracking management method, equipment and medium based on pattern recognition | |
| CN107499258A (en) | Control method and device applied to autonomous driving vehicle | |
| US20230365174A1 (en) | Automatic train tracking method and automatic train tracking system | |
| CN112859109B (en) | Unmanned aerial vehicle panoramic image processing method and device and electronic equipment | |
| CN112598668B (en) | Defect identification method and device based on three-dimensional image and electronic equipment | |
| US20200250970A1 (en) | Information processing apparatus, information processing method and program | |
| WO2021116899A1 (en) | Automotive data processing system with efficient generation and exporting of metadata | |
| US20200378774A1 (en) | Method for allocating shareable vehicles in preset areas and electronic device employing the method | |
| CN111898475A (en) | Method and device for estimating state of non-motor vehicle, storage medium, and electronic device | |
| KR101362962B1 (en) | System for recognizing and searching the car number and method therefor | |
| CN107450505A (en) | A kind of rapid detection system of vehicle, method | |
| CN205644589U (en) | License plate classification device | |
| CN108171991A (en) | A kind of identification controller and Vehicle License Plate Recognition System | |
| CN212302511U (en) | Car number and case number identification association system monitoring facilities | |
| CN106097751A (en) | Vehicle travel control method and device | |
| CN114141041A (en) | Remote parking control method, system, storage medium and device for automobile |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |