CN1296865C - Single color words reinforcing method - Google Patents
Single color words reinforcing method Download PDFInfo
- Publication number
- CN1296865C CN1296865C CNB03134769XA CN03134769A CN1296865C CN 1296865 C CN1296865 C CN 1296865C CN B03134769X A CNB03134769X A CN B03134769XA CN 03134769 A CN03134769 A CN 03134769A CN 1296865 C CN1296865 C CN 1296865C
- Authority
- CN
- China
- Prior art keywords
- point
- literal
- gray
- marginal
- monochromatic character
- 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
Landscapes
- Image Processing (AREA)
Abstract
The present invention relates to a method for reinforcing monochrome words, which is applied to digital images comprising monochrome words. The method comprises the steps: firstly, marking word points and points to be determined in an image by the gray level of pixel points; secondly, searching quasi-word points in specific points and smoothening the quasi-word points into word points; thirdly, searching word edge points in the rest points to be determined and filling the word edge points; and finally, filling the word points. Thus, the method increases the contrast of words and the background, eliminates sawtooth effect and makes the monochrome words clearer.
Description
Technical field
The present invention relates to a kind of digital image processing method, relate in particular to the method that monochromatic character strengthens in a kind of digital picture.
Background technology
Along with popularizing of digital product, digital picture more and more appears in the middle of we live, and the disposal route of grasping digital picture has become the work ability of an indispensability in the life of people.The disposal route of digital picture has a variety of, but its majority all is that graph image is handled, and big part is arranged for character image in the middle of digital picture, for the monochromatic character image, does not also have special disposal route.
There is unsharp problem in the monochromatic character image through regular meeting.For example, we are when duplicating, and the edge of black literal will be unintelligible, and the color that literal neither black.This is because the relation of ink penetration, the sharp-edged black literal of script, after overscanning, the center section of black literal is the gray scale of 60%-70%, the edge part branch of literal is the gray scale from the 60%-70% grayscale transition to 30%-40% gradually, carries out the transition to light background (＜10% gray scale) again.Therefore the edge of literal has produced the sawtooth effect, need handle character image.For graph image, algorithm by figure image intensifying and edge extracting etc. is handled image and can be made it more clear, but these disposal routes mainly all are the figures at complexity, can't be applied to the monochromatic character image, in the present various digital image processing methods, also there is not method at monochromatic character.Therefore how character image is handled,, increased the sharpness of literal, become an important topic of current digital image processing field with sawtooth effect in the removal of images.
Summary of the invention
Technical matters to be solved by this invention provides a kind of monochromatic character Enhancement Method, be applied to include the digital picture of monochromatic character, by extraction with the monochromatic character edge, and to the filling of literal point, reach the enhancing literal, eliminate the sawtooth effect, make the edge of literal more clear.
For achieving the above object, the invention provides the method that a kind of monochromatic character strengthens, be applied to the monochromatic character in the digital picture, make literal more clear, its characteristics are that this method comprises the steps: at least
According to gray values of pixel points, literal point in the marking image and point to be located;
In this point to be located, search accurate literal point, and smoothly be the literal point it;
Search the literal marginal point in the point to be located and with its filling in residue; And this literal point of filling,
Wherein, describedly search the literal marginal point in the point to be located and the step of its filling also comprised in residue:
According to the gray-scale value of point to be located, mark the target point to be located;
With this target point to be located is the center, and 8 directions of neighbor pixel are extended ray towards periphery respectively;
According to two bearing of trends of straight line, extract the gray-scale value of a plurality of continuous image vegetarian refreshments adjacent on the same straight line respectively with this target point to be located;
According to this gray-scale value, judge the affiliated type of these a plurality of continuous image vegetarian refreshments respectively;
According to this judged result, corresponding target point to be located is labeled as marginal point.
The method that above-mentioned monochromatic character strengthens, its characteristics are, also comprise the step of filling the residue point to be located with background dot.
The method that above-mentioned monochromatic character strengthens, its characteristics are that described step of searching accurate literal point in point to be located also comprises:
Extract the neighbor of point to be located;
Calculate the number of continuous literal point in this neighbor;
According to this number, determine this point to be located literal point that whether is as the criterion.
The method that above-mentioned monochromatic character strengthens, its characteristics be, and be described according to this number, determines whether it is the step of accurate literal point, as if the number of this continuous literal point greater than 5, this point to be located literal point that is as the criterion then.
The method that above-mentioned monochromatic character strengthens, its characteristics are, according to this judged result, corresponding target point to be located is labeled as the step of marginal point, if a plurality of continuous image vegetarian refreshments adjacent with this target point to be located are the literal point on the bearing of trend of this straight line, a plurality of continuous image vegetarian refreshments adjacent with this target point to be located on another bearing of trend are white point, and then this target point to be located of mark is a marginal point, wherein, described white point is the point of gray-scale value greater than a reference value.
The method that above-mentioned monochromatic character strengthens, its characteristics are, the step of the gray-scale value of a plurality of continuous image vegetarian refreshments adjacent with this target point to be located on the same straight line of described extraction is the gray-scale values that extract two adjacent with this target point to be located on same straight line continuous image vegetarian refreshments.
The method that above-mentioned monochromatic character strengthens, its characteristics are, fill the step of literal marginal point, are to fill with literal this literal marginal point of naming a person for a particular job.
The method that above-mentioned monochromatic character strengthens, its characteristics are, fill the step of literal marginal point, are to fill with literal this literal marginal point of naming a person for a particular job.
The method that above-mentioned monochromatic character strengthens, its characteristics are, fill the step of literal marginal point, are with background dot this literal marginal point to be filled.
The monochromatic character Enhancement Method of being carried according to the present invention, eliminated the sawtooth at literal edge by level and smooth, the filling of literal point, and to the extraction at edge with and to the filling of literal point and background dot, make monochromatic character and edge thereof more clear, strengthened the effect of digital picture.Disposal route of the present invention fast and has accurately improved treatment effeciency.
Describe the present invention below in conjunction with the drawings and specific embodiments, but not as a limitation of the invention.
Description of drawings
Fig. 1 is the operation workflow figure of the monochromatic character Enhancement Method that the present invention carried;
Fig. 2 is the process flow diagram of the level and smooth accurate literal point of the present invention;
Fig. 3 extracts the process flow diagram of marginal point at current point to be located for the present invention;
Fig. 4 judges the process flow diagram of continuous image vegetarian refreshments for the present invention;
Fig. 5 a～Fig. 5 d is the synoptic diagram of the level and smooth accurate literal point of the present invention;
Fig. 6 connects the synoptic diagram of the continuous image vegetarian refreshments on the bearing of trend for straight line of the present invention; And
Fig. 7 a～Fig. 7 f is the synoptic diagram of gray-value variation in the processing procedure of the present invention.
Embodiment
The present invention relates to a kind of monochromatic character Enhancement Method, be applied to include the digital picture of monochromatic character, at first by Fig. 1 method of the present invention is described, this figure is the operation workflow figure of the monochromatic character Enhancement Method that the present invention carried, and is described as follows:
At first, step 100, according to gray values of pixel points, literal point in the marking image and point to be located; Step 200 is searched accurate literal point in this point to be located, and smoothly is the literal point with it; Step 300 is searched the literal marginal point in the point to be located and with its filling in residue; Step 400 is filled this literal point.
With white background, the digital picture of black literal is an example below, and monochromatic character Enhancement Method of the present invention is elaborated.Image after overscanning, the edge of black literal is unintelligible through regular meeting, and the color that literal neither black.The present invention promptly is at this situation, makes literal more clear.According to method provided by the present invention, scan image at first obtains the gray-scale value of all pixels in the image, literal point in the marking image and point to be located.With the point of gray-scale value less than reference value T1, be labeled as the literal point according to the gray-scale value of pixel, remaining pixel is labeled as point to be located.The value of reference value T1 is 0x70 in the present embodiment, searches accurate literal point then, and level and smooth accurate literal point is the literal point.
See also Fig. 2, this figure is the process flow diagram of the level and smooth accurate literal point of the present invention.At first, step 210, the neighbor of extraction point to be located; Step 220 is calculated in this neighbor the number of literal point continuously then; Step 230, if continuously literal o'clock is greater than 5, then execution in step 240, with this point to be located mark literal point that is as the criterion.Wherein the neighbor of point to be located be meant apart from point to be located nearest around pixel on 8 directions.Then, step 250 is the literal point smoothly with accurate literal point.
Therefore accurate literal point is exactly that to have 5 continuous adjacent pixels at least be the point to be located of literal point.The accurate literal point that comprises horizontal direction, and the accurate literal point of vertical direction.Smoothing processing is also carried out in level and vertical both direction, sees also Fig. 5 a～Fig. 5 d, and it is the synoptic diagram of the level and smooth accurate literal point of the present invention.Accurate literal point 10 promptly is that 5 directions all are the point to be located of literal point on every side among the figure.Fig. 5 a and Fig. 5 b are the level and smooth synoptic diagram of horizontal direction; Fig. 5 c and Fig. 5 d are the level and smooth synoptic diagram of vertical direction.
Accurate literal point is level and smooth for behind the literal point, extraction marginal point and marginal point filled in current point to be located next.See also Fig. 3, this figure is the present invention extracts marginal point at current point to be located a process flow diagram.Step 310, scan image marks the target point to be located according to the gray-scale value of current point to be located.Described target point to be located is meant gray-scale value greater than a reference value T3 and less than the point to be located of a reference value T4, and reference value T3 and T4 are predetermined value, and the value of present embodiment reference value T3 is 0x70, and the reference value of T4 is 0xB0.Then, step 320 is the center with this target point to be located, and 8 directions of neighbor pixel are extended ray towards periphery respectively.Please in conjunction with reference to figure 6, the target point to be located is the center for a some M with target point to be located M, and 8 of neighbor pixel directions are extended 8 rays towards periphery respectively, and these 8 rays have constituted 4 straight lines, for each bar straight line L, all include two bearing of trends.Then, step 330 is extracted the gray-scale value of two adjacent with this target point to be located on each straight line continuous image vegetarian refreshments respectively.Present embodiment can certainly extract a plurality of pixels for extracting two continuous image vegetarian refreshments.With straight line L is example, is respectively continuous image vegetarian refreshments 601 and continuous image vegetarian refreshments 602 on two bearing of trends.Then, step 340 according to its gray-scale value, is judged the affiliated type of these two continuous image vegetarian refreshments respectively.
Please in conjunction with reference to figure 4, this figure is the process flow diagram that the present invention judges the continuous image vegetarian refreshments.At first, step 3410 judges whether two continuous image vegetarian refreshments are the literal point; Then, step 3420 judges whether two continuous image vegetarian refreshments are white point.If two the continuous image vegetarian refreshments adjacent with this target point to be located are literal point (gray-scale value is less than T1) on the bearing of trend of this straight line, two adjacent with this target point to be located on another bearing of trend continuous image vegetarian refreshments are white point, and then this target point to be located of mark is a marginal point.Described white point is the point of gray-scale value greater than reference value T2, T2=0xC0 in the present embodiment.
Step 350 according to this judged result, is labeled as marginal point with corresponding target point to be located.Again this marginal point is filled (step 400), it can be filled to literal point or background dot, fills this literal point at last, and the point to be located of remainder is filled with background dot, finishes the entire process process.
When carrying out above-mentioned filling, if marginal point is filled to the literal point, the literal after then filling can seem very thick, if marginal point is filled to background dot, it is very thin that literal can seem again.For the thickness degree that makes literal is moderate, the marginal point that is positioned at the left side can be filled to the literal point, the marginal point that will be positioned at the right is filled to background dot; The marginal point that perhaps will be positioned at the right is filled to the literal point, and the marginal point that is positioned at the left side is filled to background dot, can make literal thickness degree moderate, thereby the image after guaranteeing to handle does not depart from original image.
Below, seeing also Fig. 7 a～Fig. 7 f, it is the synoptic diagram of gray-value variation in the processing procedure of the present invention.Scan image at first obtains the gray-scale value of image, and Fig. 7 a is the original image after the scanning of the present invention.With the pixel of gray-scale value less than reference value T1, be labeled as the literal point then, remaining pixel is labeled as point to be located, and the image behind the mark is shown in Fig. 7 b, and red bolded section is the literal point, and remaining point is point to be located.Include literal point 10 surely among the figure, extract the accurate literal point in the image, and smoothly be the literal point it.Level and smooth result is shown in Fig. 7 c, and accurate literal point 10 has become the literal point.According to current point to be located, the edge of mark literal.See also Fig. 7 d, green bolded section is the literal edge of mark.Be literal point or background dot with edge labelling again, in the present embodiment literal marginal point be labeled as background dot.Background dot is a white point, and its gray-scale value is the point of T5 (T5=0xff), shown in Fig. 7 e, with literal marginal point filling the becoming white point of green bolded section.At last literal point is filled, literal point is a stain, and stain is meant that gray-scale value is T6 (T6=0) point.Shown in Fig. 7 f, it is 0 stain that the literal point of red runic has all become gray-scale value.Therefore literal point all is a stain, and background dot is a white point, thereby the contrast of literal and background strengthens, and has increased the sharpness of literal, has eliminated sawtooth.
Though preferred embodiment of the present invention as mentioned above; right its is not in order to limit the present invention; any those skilled in the art; without departing from the spirit and scope of the present invention; can make corresponding change and distortion, but these corresponding changes and distortion all should belong to the protection domain of the appended claim of invention.
Claims (8)
1, a kind of method of monochromatic character enhancing is applied to the monochromatic character in the digital picture, makes literal more clear, it is characterized in that this method comprises the steps: at least
According to gray values of pixel points, literal point in the marking image and point to be located;
In this point to be located, search accurate literal point, and smoothly be the literal point it;
Search the literal marginal point in the point to be located and with its filling in residue; And this literal point of filling,
Wherein, describedly search the literal marginal point in the point to be located and the step of its filling also comprised in residue:
According to the gray-scale value of point to be located, mark the target point to be located;
With this target point to be located is the center, and 8 directions of neighbor pixel are extended ray towards periphery respectively;
According to two bearing of trends of straight line, extract the gray-scale value of a plurality of continuous image vegetarian refreshments adjacent on the same straight line respectively with this target point to be located;
According to this gray-scale value, judge the affiliated type of these a plurality of continuous image vegetarian refreshments respectively;
According to this judged result, corresponding target point to be located is labeled as marginal point.
2, the method for monochromatic character enhancing according to claim 1 is characterized in that, also comprises the step of filling the residue point to be located with background dot.
3, the method for monochromatic character enhancing according to claim 1 is characterized in that described step of searching accurate literal point in point to be located also comprises:
Extract the neighbor of point to be located;
Calculate the number of continuous literal point in this neighbor;
According to this number, determine this point to be located literal point that whether is as the criterion.
4, the method that strengthens of monochromatic character according to claim 3 is characterized in that, and is described according to this number, determines whether it is the step of accurate literal point, as if the number of this continuous literal point greater than 5, this point to be located literal point that is as the criterion then.
5, the method for monochromatic character enhancing according to claim 1, it is characterized in that, according to this judged result, corresponding target point to be located is labeled as the step of marginal point, if a plurality of continuous image vegetarian refreshments adjacent with this target point to be located are the literal point on the bearing of trend of this straight line, a plurality of continuous image vegetarian refreshments adjacent with this target point to be located on another bearing of trend are white point, and then this target point to be located of mark is a marginal point, wherein, described white point is the point of gray-scale value greater than a reference value.
6, the method for monochromatic character enhancing according to claim 1, it is characterized in that, the step of the gray-scale value of a plurality of continuous image vegetarian refreshments adjacent with this target point to be located on the same straight line of described extraction is the gray-scale values that extract two adjacent with this target point to be located on same straight line continuous image vegetarian refreshments.
7, the method that strengthens of monochromatic character according to claim 1 is characterized in that, fills the step of literal marginal point, is to fill with literal this literal marginal point of naming a person for a particular job.
8, the method for monochromatic character enhancing according to claim 1 is characterized in that, fills the step of literal marginal point, is with background dot this literal marginal point to be filled.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB03134769XA CN1296865C (en) | 2003-09-29 | 2003-09-29 | Single color words reinforcing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB03134769XA CN1296865C (en) | 2003-09-29 | 2003-09-29 | Single color words reinforcing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1604135A CN1604135A (en) | 2005-04-06 |
| CN1296865C true CN1296865C (en) | 2007-01-24 |
Family
ID=34659100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB03134769XA Expired - Fee Related CN1296865C (en) | 2003-09-29 | 2003-09-29 | Single color words reinforcing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1296865C (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101291384B (en) * | 2007-04-20 | 2010-05-26 | 致伸科技股份有限公司 | Method for separating image and text and reinforcing text |
| JP4431607B2 (en) * | 2007-11-08 | 2010-03-17 | シャープ株式会社 | Image processing apparatus, image forming apparatus, image processing method, computer program, and recording medium |
| CN101431686B (en) * | 2008-12-17 | 2011-06-01 | 北大方正集团有限公司 | Method and device for converting colored plate into gray plate |
| CN102096903B (en) * | 2009-12-14 | 2013-01-02 | 北大方正集团有限公司 | Page rasterized character smooth processing method and system |
| CN102339396B (en) * | 2010-07-20 | 2014-10-01 | 北京大学 | Method and system for thickness adjustment of rasterized character |
| CN107766014B (en) * | 2017-11-06 | 2019-12-10 | 珠海奔图电子有限公司 | Character enhancement method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5796409A (en) * | 1993-04-06 | 1998-08-18 | Ecole Polytechnique Federale De Lausanne | Method for producing contrast-controlled grayscale characters |
| WO2001054064A1 (en) * | 2000-01-17 | 2001-07-26 | Koninklijke Philips Electronics N.V. | Text improvement |
| CN1400807A (en) * | 2001-07-26 | 2003-03-05 | 佳能株式会社 | Image processing method and equipment, image processing system and storage medium |
-
2003
- 2003-09-29 CN CNB03134769XA patent/CN1296865C/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5796409A (en) * | 1993-04-06 | 1998-08-18 | Ecole Polytechnique Federale De Lausanne | Method for producing contrast-controlled grayscale characters |
| WO2001054064A1 (en) * | 2000-01-17 | 2001-07-26 | Koninklijke Philips Electronics N.V. | Text improvement |
| CN1400807A (en) * | 2001-07-26 | 2003-03-05 | 佳能株式会社 | Image processing method and equipment, image processing system and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1604135A (en) | 2005-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4667062B2 (en) | Image analysis apparatus, image analysis method, and blob identification apparatus | |
| CN1275190C (en) | Method and device for correcting image askew | |
| EP2112620B1 (en) | Image binarization using dynamic sub-image division | |
| US20110158517A1 (en) | Image processing apparatus, image processing method, and computer-readable medium | |
| CN1568479A (en) | Method and apparatus for discriminating between different regions of an image | |
| EP1392047A3 (en) | Digital document processing for image enhancement | |
| CN100338618C (en) | Automatic correction method for tilted image | |
| CN1296865C (en) | Single color words reinforcing method | |
| CN1578381A (en) | Adaptive halftone scheme to preserve image smoothness and sharpness with region identification | |
| US9332154B2 (en) | Image binarization using dynamic sub-image division | |
| US7145694B2 (en) | Image processing apparatus and method | |
| CN1292593C (en) | Color image segmentation method | |
| CN1941838A (en) | File and picture binary coding method | |
| EP0870276B1 (en) | A method for transforming a gray-level image into a black-and-white image | |
| CN1310183C (en) | Binary conversion method of character and image | |
| CN1333574C (en) | Method for extracting text filed in digital image | |
| CN1797428A (en) | Method and device for self-adaptive binary state of text, and storage medium | |
| CN114267035A (en) | Document image processing method and system, electronic device and readable medium | |
| CN1694119A (en) | Method for picture binaryzation | |
| US20050105820A1 (en) | Method for enhancing monochrome text image | |
| KR100537827B1 (en) | Method for the Separation of text and Image in Scanned Documents using the Distribution of Edges | |
| CN1581926A (en) | Half-tone processing method for automatic detecting image-text | |
| Kim et al. | Detection of moving edges based on the concept of entropy and cross-entropy | |
| CN110288626B (en) | Method and apparatus for detecting text in native electronic images | |
| KR100537829B1 (en) | Method for segmenting Scan Image |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20070124 Termination date: 20130929 |