CN101894255B - Wavelet transform-based container number positioning method - Google Patents
Wavelet transform-based container number positioning method Download PDFInfo
- Publication number
- CN101894255B CN101894255B CN2010102021409A CN201010202140A CN101894255B CN 101894255 B CN101894255 B CN 101894255B CN 2010102021409 A CN2010102021409 A CN 2010102021409A CN 201010202140 A CN201010202140 A CN 201010202140A CN 101894255 B CN101894255 B CN 101894255B
- Authority
- CN
- China
- Prior art keywords
- container number
- picture
- connected region
- container
- zone
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The invention discloses a wavelet transform-based container number positioning method, and belongs to the technical field of image processing. The method comprises the following steps of: performing median filter, binarization and morphological processing on high-frequency information by using gray information of an original picture of a to-be-identified container number after one-dimensional wavelet transform, extracting connected regions by adopting a 4-neighbourhood connection method, judging the region of the container number in the original picture according to the lengths and widths of the connected regions and the position relationship between the connected regions, and finally detecting an oblique angle of the number by using Hough transform and performing rotary correction. The method realizes the positioning of all container number candidate regions by adjusting a binarized threshold value of a high-frequency coefficient so as to avoid missing positioning of the container number and improve the accurate positioning rate of the container number. The method is insensitive to noise, and small characters which may emerge on containers have no influence on the positioning result of the container numbers.
Description
Technical field
The invention belongs to image processing field, relate generally to number location technology in the container number recognition system.
Background technology
Along with rapid economy development, the handling capacity at each big harbour is increasing.Container on the harbour managed has efficiently become a problem demanding prompt solution, and this is concerning the high efficiency that container leaves the port, approaches, and also direct relation the handling capacity that can bear at harbour.And the harbour all is to use artificial form that relevant formality is registered and handled to this case for the container that passes in and out now, and this has had a strong impact on the efficient that container passes in and out, and has also influenced the carrying capacity at harbour.And the round-the-clock work of harbour needs, this guarantees that with regard to being difficult to the harbour staff does not make mistakes because of factor affecting such as fatigue, moods.In order efficiently and accurately harbour container turnover to be managed, the research of the automatic recognition system of container is just become more urgent.The container number automatic recognition system has been arranged; People just can round-the-clockly register the container of entering and leaving port automatically and handle relevant formality; Shorten the hold-up time of container thereby look up and down at the harbour; Improve container entering and leaving port efficient, can also guarantee to the registration work of container number do not receive the people factors such as mood, fatigue influence and make mistakes.
Complete container recognition system needs three steps at least: 1) container number location, 2) container number cuts apart 3) identification of container number.Because the container number arrangement mode is various and the container color background is various, this has all strengthened the difficulty of container number location.
At present existing container number positioning method mainly contains:
1, at first in picture, finds suspicious character candidates frame; Make the character candidates piece that removes non-container number as much as possible screening then; Further use of the input of the central point of character candidates piece this moment again as the hough conversion; Orient one and pass through multiword symbol candidate blocks straight line as far as possible, thereby orient the container character.
2, at first capable or column data scans to picture, is marked in delegation or the row gradient difference greater than the number of the pixel of certain threshold value, if the number of certain such pixel of row is greater than certain threshold value, then with the suspicious row of this row as container number.If do not find suspicious row in the whole picture then turn down threshold value and rescan searching.After having found the suspicious row of container number; Suspicious row is carried out further affirmation; Promptly the following continuous several line data to suspicious row carry out passing marker; When if the gauge point position of gauge point and the lastrow of each row is more or less the same then confirm that this suspicious actions container number is expert at, this has just accomplished the coarse positioning of container number.System generates queueing discipline according to existing container number arrangement mode then, and the image data that calls again behind rule and the coarse positioning compares, thereby orients correct container number plate.
Above-mentioned algorithm all has problems in the container number location to a certain extent.The algorithm of suspicious character candidates frame in the picture of location is not described in the location algorithm 1, can't be learnt the location algorithm of this character candidates frame.But will from original image, accurately orient each suspicious character candidates frame very big difficulty is arranged, its accuracy rate also can be different because of picture quality, and this just causes the accurate location rate of last container number not high.Location algorithm 2 is to utilize gray difference in the picture to carry out the coarse positioning of container number plate, and this has just determined this method comparatively responsive to noise.
Summary of the invention
The invention provides a kind of container number positioning method based on wavelet transformation.This method can be oriented container number in complex background, to insensitive for noise, the small characters that possibly occur above the container is to the not influence of container number positioning result.
Detailed technology scheme of the present invention is following:
A kind of container number positioning method based on wavelet transformation, as shown in Figure 1, may further comprise the steps:
Step 1: the former picture of container number to be identified of collection, and convert thereof into the gray scale format picture.
Because container background colour and container number do not have fixing colour match, can not locate container number according to color value, so need to convert the former picture of container number to be identified to the gray scale format picture.If the former picture of gathering of container number to be identified is the rgb format picture; The conversion formula that converts thereof into the gray scale format picture is: gray=0.229 * R+0.587 * G+0.114 * B; Wherein gray representes certain gray values of pixel points in the gray scale format picture, and R, G and B represent the pixel value of redness, green and blue three passages of this pixel in the rgb format picture respectively.
Step 2: select the harr small echo for use, the gray scale format picture of step 1 gained is carried out one-dimensional wavelet transform.
Step 3: the high frequency coefficient that obtains behind the wavelet transformation is formed picture, this picture is carried out medium filtering.
Medium filtering is the influence for cancelling noise, and it is removed the isolated noise in the picture has good effect.The window of medium filtering is 3 * 3 pixels or 5 * 5 pixel sizes.
Step 4: the picture behind the medium filtering is carried out binary conversion treatment.
Because the high frequency coefficient 90% behind the medium filtering concentrates on below 30, so the binary-state threshold during binary conversion treatment can be confirmed between 3-5.
Step 5: the binaryzation picture is carried out morphology handle.
The structure that employing is handled as morphology with respect to half big or small structure of container number code character corrodes the advanced row of binaryzation picture and then expands.
Because the place's greyscale transformation of container number code character is violent, has bigger high frequency coefficient behind the wavelet transformation.After the binaryzation; The container number place is all turned to 1 by two-value; Other background places are then turned to 0 by two-value, but might occur exist behind some or several character binaryzations with other characters the time do not couple together, also the partial noise zone can appear.Though such noise region by two-value turn to 1 should the zone less, and by isolated far.The use morphological method is corroded this picture earlier and then is expanded, and can the container number that noise region is rejected and the while just ruptures be coupled together.
Step 6: extract morphology and handle the connected region in the binaryzation picture of back.
Adopt 4-neighborhood communicating method to mark morphology and handle the connected region in the binaryzation picture of back, and count size, the positional information of each connected region one by one.
Step 7: the connected region to step 6 extracts is judged, and is found container number The corresponding area in former picture.
The container number arrangement mode comprises row, two row, delegation and two row.Because there is diversity in the arrangement mode of container number itself, so need carry out analysis and judgement to each connected region, which connected region interpretation goes out is only real number zone.During practical operation; Position according between length, width and the connected region of connected region concerns the zone of judging container number in the former picture: at first choose the longest connected region; Judge whether this zone reaches the container number code length of a delegation or a column number arrangement mode, if then this longest connected region The corresponding area in former picture is exactly the container number zone; If not, then the longest connected region adds that second long connected region The corresponding area in former picture is exactly the container number zone.
Step 8: detect the degree of tilt in step 7 gained container number zone, and do corresponding rotation adjustment.
At first extract step 7 gained container number edges of regions point, utilize the degree of tilt in Hough (Hough) change detection container number zone then,, then utilize interpolation algorithm that it is rotated to level if degree of tilt surpasses 2 °.
Through above step, just can in the former picture of container number to be identified, orient the number region.Certainly also just the container number coarse positioning is come out through after the above step, each character accurately located and be partitioned into to the identification that arrive container number also need to the container number plate that comes out in the location.What the present invention adopted is the character zone that wavelet transformation is sought container; Because character zone greyscale transformation is bigger; High frequency coefficient after its conversion is just bigger; So the present invention utilizes high frequency coefficient just can orient the container number region, morphology is wherein handled the problem that also can be good at solving the picture noise.Thereby the present invention can adjust the high frequency coefficient binary-state threshold and realize orienting all container number candidate regions, thereby avoids the leakage location of container number, improves the accurate location rate of container number.
Description of drawings
Fig. 1 is a schematic flow sheet of the present invention.
Embodiment
A kind of container number positioning method based on wavelet transformation, as shown in Figure 1, may further comprise the steps:
Step 1: the former picture of container number to be identified of collection, and convert thereof into the gray scale format picture.
Because container background colour and container number do not have fixing colour match, can not locate container number according to color value, so need to convert the former picture of container number to be identified to the gray scale format picture.If the former picture of gathering of container number to be identified is the rgb format picture; The conversion formula that converts thereof into the gray scale format picture is: gray=0.229 * R+0.587 * G+0.114 * B; Wherein gray representes certain gray values of pixel points in the gray scale format picture, and R, G and B represent the pixel value of redness, green and blue three passages of this pixel in the rgb format picture respectively.
Step 2: select the harr small echo for use, the gray scale format picture of step 1 gained is carried out one-dimensional wavelet transform.
Step 3: the high frequency coefficient that obtains behind the wavelet transformation is formed picture, this picture is carried out medium filtering.
Medium filtering is the influence for cancelling noise, and it is removed the isolated noise in the picture has good effect.The window of medium filtering is 3 * 3 pixels or 5 * 5 pixel sizes.
Step 4: the picture behind the medium filtering is carried out binary conversion treatment.
Because the high frequency coefficient 90% behind the medium filtering concentrates on below 30, so the binary-state threshold during binary conversion treatment can be confirmed between 3-5.
Step 5: the binaryzation picture is carried out morphology handle.
The structure that employing is handled as morphology with respect to half big or small structure of container number code character corrodes the advanced row of binaryzation picture and then expands.
Because the place's greyscale transformation of container number code character is violent, has bigger high frequency coefficient behind the wavelet transformation.After the binaryzation; The container number place is all turned to 1 by two-value; Other background places are then turned to 0 by two-value, but might occur exist behind some or several character binaryzations with other characters the time do not couple together, also the partial noise zone can appear.Though such noise region by two-value turn to 1 should the zone less, and by isolated far.The use morphological method is corroded this picture earlier and then is expanded, and can the container number that noise region is rejected and the while just ruptures be coupled together.
Step 6: extract morphology and handle the connected region in the binaryzation picture of back.
Adopt 4-neighborhood communicating method to mark morphology and handle the connected region in the binaryzation picture of back, and count size, the positional information of each connected region one by one.
Step 7: the connected region to step 6 extracts is judged, and is found container number The corresponding area in former picture.
The container number arrangement mode comprises row, two row, delegation and two row.Because there is diversity in the arrangement mode of container number itself, so need carry out analysis and judgement to each connected region, which connected region interpretation goes out is only real number zone.During practical operation; Position according between length, width and the connected region of connected region concerns the zone of judging container number in the former picture: at first choose the longest connected region; Judge whether this zone reaches the container number code length of a delegation or a column number arrangement mode, if then this longest connected region The corresponding area in former picture is exactly the container number zone; If not, then the longest connected region adds that second long connected region The corresponding area in former picture is exactly the container number zone.
Step 8: detect the degree of tilt in step 7 gained container number zone, and do corresponding rotation adjustment.
At first extract step 7 gained container number edges of regions point, utilize the degree of tilt in Hough (Hough) change detection container number zone then,, then utilize interpolation algorithm that it is rotated to level if degree of tilt surpasses 2 °.
Through above step, just can in the former picture of container number to be identified, orient the number region.
Claims (2)
1. container number positioning method based on wavelet transformation may further comprise the steps:
Step 1: the former picture of container number to be identified of collection, and convert thereof into the gray scale format picture; If the former picture of gathering of container number to be identified is the rgb format picture; The conversion formula that converts thereof into the gray scale format picture is: gray=0.229 * R+0.587 * G+0.114 * B; Wherein gray representes certain gray values of pixel points in the gray scale format picture, and R, G and B represent the pixel value of redness, green and blue three passages of this pixel in the rgb format picture respectively;
Step 2: select the harr small echo for use, the gray scale format picture of step 1 gained is carried out one-dimensional wavelet transform;
Step 3: the high frequency coefficient that obtains behind the wavelet transformation is formed picture, this picture is carried out medium filtering;
Step 4: the picture behind the medium filtering is carried out binary conversion treatment; Binary-state threshold during binary conversion treatment is confirmed between 3-5;
Step 5: the binaryzation picture is carried out morphology handle;
The structure that employing is handled as morphology with respect to half big or small structure of container number code character corrodes the advanced row of binaryzation picture and then expands;
Step 6: extract morphology and handle the connected region in the binaryzation picture of back;
Adopt 4-neighborhood communicating method to mark morphology and handle the connected region in the binaryzation picture of back, and count size, the positional information of each connected region one by one;
Step 7: the connected region to step 6 extracts is judged, and is found container number The corresponding area in former picture;
Position according between length, width and the connected region of connected region concerns the zone of judging container number in the former picture: at first choose the longest connected region; Judge whether this zone reaches the container number code length of a delegation or a column number arrangement mode, if then this longest connected region The corresponding area in former picture is exactly the container number zone; If not, then the longest connected region adds that second long connected region The corresponding area in former picture is exactly the container number zone;
Step 8: detect the degree of tilt in step 7 gained container number zone, and do corresponding rotation adjustment;
At first extract step 7 gained container number edges of regions point, utilize Hough transformation to detect the degree of tilt in container number zone then,, then utilize interpolation algorithm that it is rotated to level if degree of tilt surpasses 2 °.
2. the container number positioning method based on wavelet transformation according to claim 1 is characterized in that, the window of step 3 medium filtering is 3 * 3 pixels or 5 * 5 pixel sizes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102021409A CN101894255B (en) | 2010-06-13 | 2010-06-13 | Wavelet transform-based container number positioning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102021409A CN101894255B (en) | 2010-06-13 | 2010-06-13 | Wavelet transform-based container number positioning method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101894255A CN101894255A (en) | 2010-11-24 |
| CN101894255B true CN101894255B (en) | 2012-04-25 |
Family
ID=43103443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102021409A Expired - Fee Related CN101894255B (en) | 2010-06-13 | 2010-06-13 | Wavelet transform-based container number positioning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101894255B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102184398B (en) * | 2011-06-17 | 2012-07-18 | 电子科技大学 | Edge-detection-based container number positioning method |
| CN102878044B (en) * | 2012-10-17 | 2015-01-07 | 飞翼股份有限公司 | Full hydraulic control mining concrete pump |
| CN110598697A (en) * | 2019-08-23 | 2019-12-20 | 上海撬动网络科技有限公司 | Container number positioning method based on thickness character positioning |
| CN116452467B (en) * | 2023-06-16 | 2023-09-22 | 山东曙岳车辆有限公司 | Container real-time positioning method based on laser data |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1343946A (en) * | 2000-09-14 | 2002-04-10 | 伊斯曼柯达公司 | Parallel reversely dispersed small wave change |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8503809B2 (en) * | 2007-05-17 | 2013-08-06 | Sony Corporation | Information processing apparatus and method to entropy code upon processing of predetermined number of precincts |
-
2010
- 2010-06-13 CN CN2010102021409A patent/CN101894255B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1343946A (en) * | 2000-09-14 | 2002-04-10 | 伊斯曼柯达公司 | Parallel reversely dispersed small wave change |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101894255A (en) | 2010-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109271991B (en) | License plate detection method based on deep learning | |
| KR100339691B1 (en) | Apparatus for recognizing code and method therefor | |
| CN107610114B (en) | optical satellite remote sensing image cloud and snow fog detection method based on support vector machine | |
| US9396404B2 (en) | Robust industrial optical character recognition | |
| CN103235938B (en) | The method and system of car plate detection and indentification | |
| CN101706873B (en) | Identification method and device of digital-class limitation marking | |
| CN107491730A (en) | A kind of laboratory test report recognition methods based on image procossing | |
| US20070253040A1 (en) | Color scanning to enhance bitonal image | |
| US9977957B2 (en) | Method and system for identifying characters on a valuable medium | |
| CN109409355B (en) | Novel transformer nameplate identification method and device | |
| CN109670515A (en) | A kind of detection method and system changed for building in unmanned plane image | |
| CN102426649A (en) | Simple steel seal digital automatic identification method with high accuracy rate | |
| CN102750541A (en) | Document image classifying distinguishing method and device | |
| CN104680161A (en) | Digit recognition method for identification cards | |
| CN107195069A (en) | A kind of RMB crown word number automatic identifying method | |
| CN101894255B (en) | Wavelet transform-based container number positioning method | |
| CN108073849B (en) | Bar code detection method, device and system | |
| Liu et al. | A novel multi-oriented chinese text extraction approach from videos | |
| CN103593695A (en) | Method for positioning DPM two-dimension code area | |
| CN105976407B (en) | A kind of service plate shape recognition process suitable under obstruction conditions | |
| CN105303363A (en) | Data processing method and data processing system | |
| Zhang et al. | A combined algorithm for video text extraction | |
| CN108171098A (en) | A kind of code detection method and equipment | |
| CN110298835B (en) | Leather surface damage detection method, system and related device | |
| CN103886319A (en) | Intelligent held board recognizing method based on machine vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120425 Termination date: 20180613 |