CN104657731A - Container number correction method - Google Patents
Container number correction method Download PDFInfo
- Publication number
- CN104657731A CN104657731A CN201510098184.4A CN201510098184A CN104657731A CN 104657731 A CN104657731 A CN 104657731A CN 201510098184 A CN201510098184 A CN 201510098184A CN 104657731 A CN104657731 A CN 104657731A
- Authority
- CN
- China
- Prior art keywords
- corrected
- single character
- container
- antidote
- container number
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/243—Aligning, centring, orientation detection or correction of the image by compensating for image skew or non-uniform image deformations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Abstract
The invention discloses a container number correction method and relates to the technical field of optical character recognition and correction. Acquired container numbers can be corrected with the correction method, so that character deformation caused by the shooting angle factor, the container inclination factor and the like is eliminated, and the container number recognition rate is increased finally. The container number correction method specifically comprises multiple steps as follows: performing image preprocessing and character segmentation on the container numbers, repeatedly and continuously rotating to-be-corrected single-character segmentation graphs, calculating corresponding minimum enclosing rectangles in each rotating process, comparing height-to-width ratios and the like.
Description
Technical field
The present invention relates to optical character identification and correcting technology field, particularly relate to a kind of antidote of container number.
Background technology
Container ship maximizes increasingly on the one hand, and container liner has again stricter sailing date requirement on the other hand, for ensureing the punctuality rate of sailing schedule, proposes the requirement that container wharf needs to complete as soon as possible the loading and unloading operation work of container cargo.For this reason, how to utilize the modernization means such as robot calculator and data exchange system, the information system management level improving container wharf becomes the important subject in container technology field.China is at present except development level is in Hong Kong HIT in prostatitis, outside Shanghai SCT and container wharf, salt pan etc., most container wharf computer management level is not high, usually there is the job instruction that computing machine assigns and can not arrive executor in time, the information that executor is finished can not return computing machine in time, and produce data and " overstock " phenomenon, have impact on harbour throughput rate.The basic way addressed this problem is the complete computerizing management realizing container wharf as early as possible.At present, yard management and the management of handling ship of the many container wharfs of China realize computerize all, and place's container number identifications such as gate, container wharf management rely on the identification of human eye, the mode of operation of hand-kept completely, therefore labor strength is large, work efficiency is lower.
Research container number Intelligent Recognition administrating system is one of most important and the most difficult link realizing the management of harbour service complete computerizing.Container number identifies that administrating system utilizes computer system to complete case number (CN) identification and record expeditiously automatically, overcomes the various unfavorable factors that manual operation brings, raises labour efficiency.At gate, container wharf, container number detection system is set, when container is through gate detection system, system automatically identifies container number, feeds back to computer system, computer system is by after signal transacting, and just bootable driver completes loading and unloading to the container of specifying.
Container number uniquely identify each container in global range, and therefore, the automatic identification of container just can be summed up as the identification to case number (CN), as long as case number (CN) determines, other information about this container also just determine.Container automatic recognition system based on optics case number (CN) recognition technology identifies container by identifying to brush at the case number (CN) of tank surface.Concrete, container number by 4 capitalization English letters and 7 arabic numeral totally 11 characters form.Wherein 4 letters are divided into case main code, EIC equipment identification code two parts, and 7 numerals are divided into sequence number, check code two parts.
In container number recognition system, the segmentation figure of single character can be obtained by Image semantic classification, case number (CN) location and Character segmentation, after normalization and pre-service, just can enter case number (CN) cognitive phase.But inventor finds, due to problems such as shooting environmental, angle and casing inclinations, case number (CN) figure can be made with very large complicacy.The single Character segmentation figure obtained in these complex situations, tilt quantity wherein will certainly impact recognition result, may cause identifying an error result under serious conditions.Therefore, be necessary to carry out correction process to the segmentation figure of the single character tilted before carrying out case number (CN) identification.
Summary of the invention
The invention provides a kind of antidote of container number, this antidote can be corrected the case number (CN) obtained, thus eliminates the Character deformation situation because the factors such as shooting angle, casing inclination cause, the final discrimination improving container number.
For solving the problems of the technologies described above, present invention employs following technical scheme:
An antidote for container number, comprises the steps:
S1: to container number carry out Image semantic classification go forward side by side line character separate, form single Character segmentation figure to be corrected;
S2: single character picture to be corrected is rotated the first angle in the counterclockwise direction; Then, with the fixed angle single character picture that continuous rotation is to be corrected along clockwise direction, until single character picture to be corrected turns over the second angle along clockwise direction;
S3: calculate the minimum enclosed rectangle that single character picture to be corrected in each rotary course is corresponding, and the height-width ratio calculating each minimum enclosed rectangle respectively;
S4: comparative statistics obtains the minimum enclosed rectangle with maximum height width ratio, this rotational angle having the minimum enclosed rectangle of maximum height width ratio corresponding is the RA needed for single character picture to be corrected.
Preferred further, described antidote also comprises after step S1, before step S2,
Step S11: single Character segmentation figure normalized to be corrected is become bianry image.
Comparatively preferred, described first angle is 30 ° to 45 °.
Comparatively preferred, described second angle is 60 ° to 90 °.
Comparatively preferred, described fixed angle is 2.5 ° to 5 °.
Concrete, the minimum enclosed rectangle that single character picture to be corrected is corresponding is specially: with maximum horizontal ordinate in single character picture to be corrected, minimum horizontal ordinate, maximum ordinate, minimum ordinate for determining the rectangle that formed in border.
The invention provides a kind of antidote of container number, wherein this antidote includes Image semantic classification and Character segmentation, repeatedly continuous rotation single Character segmentation figure to be corrected, calculates minimum enclosed rectangle corresponding in each rotary course, compares some steps such as height-width ratio; By above-mentioned steps, this antidote can be corrected the container number of run-off the straight, thus improves the success ratio identifying container number.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the antidote of a kind of container number of the present invention;
Fig. 2 is the schematic diagram that antidote step S2 many continuous rotations of a kind of container number of the present invention single character partition graph to be corrected becomes;
Fig. 3 be the minimum enclosed rectangle that single Character segmentation figure that the antidote step S3 of a kind of container number of the present invention is to be corrected is corresponding height-width than and the contrast relationship table of the anglec of rotation.
Embodiment
The invention provides a kind of antidote of container number, this antidote can be corrected the case number (CN) obtained, thus eliminates the Character deformation situation because the factors such as shooting angle, casing inclination cause, the final discrimination improving container number.
Below in conjunction with following accompanying drawing, the embodiment of the present invention is described in detail.
The invention provides a kind of antidote of container number, as shown in Figure 1, this antidote comprises the following steps:
Step S1: to container number carry out Image semantic classification go forward side by side line character separate, form single Character segmentation figure to be corrected.
Concrete, due to the complicacy of container and the difference of case number (CN) arrangement mode, before carrying out case number (CN) identification and correcting, first need to carry out Image semantic classification to container number region.The object of preprocessing process is: remove the noise class (such as: the rust staining on container body, spot, paint stain etc.) on container and the image-region that retains with case number (CN) character, thus alleviates the impact that noise class may bring subsequent step.
Then character separation is carried out to container number image, the string segmentation on case number (CN) is become several segmentation figure only with single character (single character specifically includes English character, numerical character etc.), thus form single Character segmentation figure to be corrected.Concrete, projection histogram method such as can be adopted to split (namely carrying out the container number region after Image semantic classification) original sample.First, projection histogram division is carried out to original sample, and calculate the Wave crest and wave trough projection relation presented between the number of pixels of character and inter-character space in histogram and the number of pixels of character.Then, analyze above-mentioned projection histogram, choose in histogram be highly less than a certain threshold value " low ebb " as intercharacter cut-point, and ensure that the segmentation spacing of horizontal case number (CN) is greater than the half of case number (CN) height, the segmentation spacing of longitudinal case number (CN) is greater than the width of case number (CN).
It should be noted that and adopt projection histogram separating character to be only applicable to case number (CN) original sample comparatively clearly; And those skilled in the art can design according to actual needs and select suitable dividing method to obtain single character separation figure to be corrected.
Step S2: single character picture to be corrected is rotated the first angle in the counterclockwise direction; Then, with the fixed angle single character picture that continuous rotation is to be corrected along clockwise direction, until single character picture to be corrected turns over the second angle along clockwise direction.
On the basis of completing steps S1, the single character picture treating rectification carries out rotation process.Concrete, the single character picture first treating rectification is rotated counterclockwise, and single character picture to be corrected is turned over the first angle in the counterclockwise direction, and wherein this first angle is chosen as 30 ° to 45 °.
Then the single character picture treating rectification carries out repeatedly continuous print and rotates clockwise, with a fixed angle single character picture that continuous rotation is to be corrected along clockwise direction, until single character picture to be corrected turns over the second angle along clockwise direction, wherein fixed angle is preferably 2.5 ° to 5 °, and the second angle is 60 ° to 90 °.
For example, using case number (CN) " 8 " as single character picture to be corrected, first this single character picture is rotated counterclockwise 30 °, then as fixed angle, it is repeatedly rotated clockwise continuously using 3 °, until this single character image clockwise rotates 60 °.With reference to (a) in figure 2, Fig. 2, the image that wherein rotation process is formed can represent that carrying out continuous 5 fixed angles to single character picture rotates the rear image (namely after rotating counterclockwise 30 °, have passed through again and rotate clockwise 15 °) formed; In Fig. 2, (b) represents that carrying out continuous 10 fixed angles to single character picture rotates the rear image formed; In Fig. 2, (c) represents that carrying out continuous 15 fixed angles to single character picture rotates the rear image formed.
You need to add is that, as a kind of preferred implementation of the present invention, described antidote also comprises after step S1, before step S2, step S11: single Character segmentation figure normalized to be corrected is become bianry image.Wherein, normalized can eliminate the unnecessary color in single Character segmentation figure to be corrected, thus improves identification rectification precision further.
Step S3: calculate the minimum enclosed rectangle that single character picture to be corrected in each rotary course is corresponding, and the height-width ratio calculating each minimum enclosed rectangle respectively.
On the basis of completing steps S2, calculate the minimum enclosed rectangle that single character picture to be corrected in each rotary course is corresponding.Concrete, the minimum enclosed rectangle that single character picture to be corrected is corresponding is maximum horizontal ordinate in single character picture to be corrected, minimum horizontal ordinate, maximum ordinate, minimum ordinate are the rectangle that border is determined to be formed.
Determining after the minimum enclosed rectangle that the single character picture corrected is corresponding, calculating the height-width ratio of each minimum enclosed rectangle further.As shown in Figure 3, wherein, the transverse axis of Fig. 3 is the rotational angle of single character picture to be corrected, and the longitudinal axis is the height-width ratio of the minimum enclosed rectangle of the correspondence of single character picture to be corrected in rotation process.
Step S4: comparative statistics obtains the minimum enclosed rectangle with maximum height width ratio, this rotational angle having the minimum enclosed rectangle of maximum height width ratio corresponding is the RA needed for single character picture to be corrected.
On the basis of completing steps S3, analyze the height-width ratio of minimum enclosed rectangle, thus determine the angle that house of correction needs.Wherein, the minimum enclosed rectangle with maximum height width ratio can be determined by Fig. 3, and determine the rotational angle corresponding to it according to this minimum enclosed rectangle further, in Fig. 3, this rotational angle is counterclockwise 15 °, and this rotational angle is the RA corrected shown in Fig. 2 needed for single character picture.
So far, this antidote obtains the RA corrected needed for single character picture by above-mentioned steps.Then, those skilled in the art can identify the single character picture after rectification further, and therefore not to repeat here.
The invention provides a kind of antidote of container number, wherein this antidote includes Image semantic classification and Character segmentation, repeatedly continuous rotation single Character segmentation figure to be corrected, calculates minimum enclosed rectangle corresponding in each rotary course, compares some steps such as height-width ratio; By above-mentioned steps, this antidote can be corrected the container number of run-off the straight, thus improves the success ratio identifying container number.
The above; be only the specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; change can be expected easily or replace, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of described claim.
Claims (6)
1. an antidote for container number, is characterized in that, comprises the steps:
S1: to container number carry out Image semantic classification go forward side by side line character separate, form single Character segmentation figure to be corrected;
S2: single character picture to be corrected is rotated the first angle in the counterclockwise direction; Then, with the fixed angle single character picture that continuous rotation is to be corrected along clockwise direction, until single character picture to be corrected turns over the second angle along clockwise direction;
S3: calculate the minimum enclosed rectangle that single character picture to be corrected in each rotary course is corresponding, and the height-width ratio calculating each minimum enclosed rectangle respectively;
S4: comparative statistics obtains the minimum enclosed rectangle with maximum height width ratio, this rotational angle having the minimum enclosed rectangle of maximum height width ratio corresponding is the RA needed for single character picture to be corrected.
2. the antidote of container number according to claim 1, is characterized in that, described antidote also comprises after step S1, before step S2,
Step S11: single Character segmentation figure normalized to be corrected is become bianry image.
3. the antidote of container number according to claim 1 and 2, is characterized in that, described first angle is 30 ° to 45 °.
4. the antidote of container number according to claim 1 and 2, is characterized in that, described second angle is 60 ° to 90 °.
5. the antidote of container number according to claim 1 and 2, is characterized in that, described fixed angle is 2.5 ° to 5 °.
6. the antidote of container number according to claim 1 and 2, it is characterized in that, the minimum enclosed rectangle that single character picture to be corrected is corresponding is specially: with maximum horizontal ordinate in single character picture to be corrected, minimum horizontal ordinate, maximum ordinate, minimum ordinate for determining the rectangle that formed in border.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510098184.4A CN104657731A (en) | 2015-03-06 | 2015-03-06 | Container number correction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510098184.4A CN104657731A (en) | 2015-03-06 | 2015-03-06 | Container number correction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104657731A true CN104657731A (en) | 2015-05-27 |
Family
ID=53248831
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510098184.4A Pending CN104657731A (en) | 2015-03-06 | 2015-03-06 | Container number correction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104657731A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106407979A (en) * | 2016-10-25 | 2017-02-15 | 深圳怡化电脑股份有限公司 | Bill character correction method and device |
| CN106446893A (en) * | 2016-09-14 | 2017-02-22 | 苏州佳世达电通有限公司 | Tilt correction method and tilt correction system for license plate recognition |
| CN109784331A (en) * | 2019-01-08 | 2019-05-21 | 河北科技大学 | Bar section tagging scheme and character picture antidote based on index point |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101246551A (en) * | 2008-03-07 | 2008-08-20 | 北京航空航天大学 | Fast license plate locating method |
| CN101751785A (en) * | 2010-01-12 | 2010-06-23 | 杭州电子科技大学 | Automatic license plate recognition method based on image processing |
| CN103455815A (en) * | 2013-08-27 | 2013-12-18 | 电子科技大学 | Self-adaptive license plate character segmentation method in complex scene |
-
2015
- 2015-03-06 CN CN201510098184.4A patent/CN104657731A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101246551A (en) * | 2008-03-07 | 2008-08-20 | 北京航空航天大学 | Fast license plate locating method |
| CN101751785A (en) * | 2010-01-12 | 2010-06-23 | 杭州电子科技大学 | Automatic license plate recognition method based on image processing |
| CN103455815A (en) * | 2013-08-27 | 2013-12-18 | 电子科技大学 | Self-adaptive license plate character segmentation method in complex scene |
Non-Patent Citations (1)
| Title |
|---|
| 洪沛霖: "手写体数字识别方法的研究与实现", 《合肥工业大学硕士学位论文》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446893A (en) * | 2016-09-14 | 2017-02-22 | 苏州佳世达电通有限公司 | Tilt correction method and tilt correction system for license plate recognition |
| CN106407979A (en) * | 2016-10-25 | 2017-02-15 | 深圳怡化电脑股份有限公司 | Bill character correction method and device |
| CN106407979B (en) * | 2016-10-25 | 2019-12-10 | 深圳怡化电脑股份有限公司 | Method and device for correcting bill characters |
| CN109784331A (en) * | 2019-01-08 | 2019-05-21 | 河北科技大学 | Bar section tagging scheme and character picture antidote based on index point |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| RU2678485C1 (en) | Method of character segmentation and recognition | |
| US20190171862A1 (en) | Method of extracting image of port wharf through multispectral interpretation | |
| EP1678659B1 (en) | Method and image processing device for analyzing an object contour image, method and image processing device for detecting an object, industrial vision apparatus, smart camera, image display, security system, and computer program product | |
| EP3330885A1 (en) | Decoding method and system for use when qr code position detection graph is damaged | |
| CN105069389A (en) | Two-dimensional code partitioning decoding method and system | |
| CN102693431B (en) | Method and device for identifying type of white number plate | |
| CN101853333A (en) | Method for picking marks in medical robot navigation positioning images | |
| CN104657731A (en) | Container number correction method | |
| CN104517108A (en) | Method and system for determining edge line of QR code binarized image | |
| CN105069452A (en) | Straight line removing method based on local structure analysis | |
| WO2020107326A1 (en) | Lane line detection method, device and computer readale storage medium | |
| CN111598952A (en) | Multi-scale cooperative target design and online detection and identification method and system | |
| CN104680109A (en) | Image recognition-based location method for bar code area | |
| CN106447615A (en) | Tilt correction method for scanning document image | |
| CN113158895B (en) | Bill identification method and device, electronic equipment and storage medium | |
| CN110674812A (en) | Civil license plate positioning and character segmentation method facing complex background | |
| CN108734161B (en) | Method, device and equipment for identifying prefix number area and storage medium | |
| US7072496B2 (en) | Slap print segmentation system and method | |
| Wang et al. | Road detection based on illuminant invariance and quadratic estimation | |
| CN105354575A (en) | Image binarization threshold determination method in sea-sky-line detection | |
| CN109118494B (en) | Overlapping region segmentation method and device based on concave point matching | |
| CN112200191B (en) | Image processing method, image processing device, computing equipment and medium | |
| CN109359497A (en) | A kind of positioning and recognition methods in VAT invoice two-dimensional image code region | |
| CN109492639A (en) | " loaded " position three-dimensional coordinate acquisition methods, system and image recognition apparatus | |
| CN105574860A (en) | Method for identifying deflection polarity of rudder slice |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150527 |