CN102289668A - Binaryzation processing method of self-adaption word image based on pixel neighborhood feature - Google Patents
Binaryzation processing method of self-adaption word image based on pixel neighborhood feature Download PDFInfo
- Publication number
- CN102289668A CN102289668A CN2011102641444A CN201110264144A CN102289668A CN 102289668 A CN102289668 A CN 102289668A CN 2011102641444 A CN2011102641444 A CN 2011102641444A CN 201110264144 A CN201110264144 A CN 201110264144A CN 102289668 A CN102289668 A CN 102289668A
- Authority
- CN
- China
- Prior art keywords
- mrow
- value
- point
- text image
- msub
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 41
- 238000012545 processing Methods 0.000 claims abstract description 20
- 238000004364 calculation method Methods 0.000 claims description 19
- 238000011835 investigation Methods 0.000 claims description 11
- 230000003044 adaptive effect Effects 0.000 claims description 8
- 230000006835 compression Effects 0.000 claims description 8
- 238000007906 compression Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 5
- 238000012015 optical character recognition Methods 0.000 description 3
- 238000005452 bending Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
Images
Landscapes
- Character Input (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method of carrying out binaryzation processing on a self-adaption word image based on a pixel neighborhood feature through a global threadhold value and a local threadhold value. At first, the brightness of a text image is adjusted globally, the proportion of a darker part in the image is increased, and the grey contrast of an object character and a background is improved in the text image; secondly, a bicubic interpolation algorithm is adopted to zoom the text image and gaps among character strokes are increased; thirdly, the stroke width d of the object character is numbered and consequently the size w of a neighborhood computation template is determined; fourthly, word information of the text image is divided into word blocks according to the determined size of the neighborhood computation template; and finally the global threadhold value and the local threadhold value are combined so that the binaryzation processing is carried out on each word block point to point. Through the binaryzation processing method, the character strokes can be separated from the background effectively; the phenomena of omission and artifact are avoided; the connectivity of the strokes is kept; the computation of the local threadhold value of pixel points in the text image is reduced, and the computation speed is greatly improved.
Description
Technical Field
The invention relates to a binarization processing method of a character image, in particular to a binarization processing method of an adaptive character image based on pixel neighborhood characteristics by utilizing global and local thresholds.
Background
Binarization is a basic technique in digital image processing technology, and is also a preprocessing technique of many image processing techniques, and is widely used in image processing such as automatic object recognition (ATR), image analysis, text enhancement, and Optical Character Recognition (OCR). Most of the existing binarization methods belong to thresholding methods, and in different applications, the selection of a threshold value determines the retention of image characteristic information. Therefore, the automatic threshold value selecting method is very worthy of study, and a good automatic threshold value selecting method not only can keep useful information in an image, but also can reduce the time overhead.
The key of the image binarization technology lies in how to select a threshold value, and the threshold value is mainly divided into three categories according to the processing modes of the threshold value on pixels:
(1) global thresholding: the whole image is binarized by using a single threshold value T (global threshold value). A global threshold T is typically determined from the histogram or spatial distribution of the gray levels of the image, and the gray level of each pixel in the image is compared to T. If the value is larger than T, the foreground color is selected; otherwise, the color is taken as the background color. Typical global thresholding methods are the Ostu method, the maximum entropy method, and the like. The global threshold method has a prominent effect when the gray scale of the target and the background is obviously different, but the method often ignores details easily, and when more shadows exist in the image or the gray scale change of the image is complex, the ideal effect is often difficult to obtain.
(2) Local thresholding: the threshold of the pixel is determined by the gray value of the current pixel and the gray characteristic of the points around the pixel. And comparing the gray level of the investigation point with the neighborhood point by defining the neighborhood of the investigation point and using a neighborhood calculation template. Typical local thresholding methods are the Bernsen method, the nillblack method, and the like. The local threshold method can adapt to more complex conditions and is more widely applied than the global threshold method. However, the boundary characteristic information of the image is often ignored, so that different areas in the original image become large areas after binarization, and some important information of the binarization result image is lost. In some applications, such as medical image segmentation and particle analysis, the result image is often required to better retain the boundary feature information, which is very important for subsequent image analysis.
(3) Dynamic threshold method: when the illumination is not uniform or the background gray scale changes greatly, different thresholds must be automatically determined according to the coordinate position relationship of the pixels, and dynamic threshold determination is implemented. The threshold selection of this method depends not only on the gray values of the pixel and surrounding pixels, but also on the coordinate position of the pixel. The dynamic threshold binarization can process an image with poor quality and even a unimodal histogram, but because the dynamic thresholding method usually needs to calculate a threshold value for each pixel point in the image, that is, the calculation amount for calculating a threshold surface (usually a curved surface) for the whole image is very large, the calculation speed is generally slow, and the development of the method is hindered to a certain extent due to the defects of time consumption and certain distortion. Iterative methods are a more common dynamic threshold determination technique.
Disclosure of Invention
In view of the above disadvantages, the present invention provides a binarization processing method for an adaptive text image based on pixel neighborhood characteristics by using global and local thresholds, comprising:
a) carrying out global brightness adjustment on the text image, and improving the gray scale contrast of the target character and the background in the text image;
b) adaptively selecting the size of a neighborhood calculation template;
c) dividing the text information of the text image into text blocks according to the size of the selected domain calculation template;
d) and performing point-by-point binarization processing on each character block by adopting a method of combining global and local threshold values.
The step a) and the step b) are also provided with:
and ab) carrying out zooming processing on the text image by adopting a bicubic interpolation algorithm.
The step a) comprises the following steps:
a1) defining the pixel value of any point of the text image as the brightness value Y of the point, normalizing the brightness value Y, I represents the brightness value of the point after normalization,
I＝Y/255；
a2) calculating the average value of the gray scale of the whole imageal，
Wherein rho epsilon I represents that rho points are in a definition domain of the text image, and N represents the number of pixel values in the text image;
a3) a global luminance dynamic range compression degree coefficient gamma is defined,
a4) global brightness adjustment is carried out on the text image by utilizing the global brightness dynamic range compression degree coefficient gamma,
I′＝Iγ；
a5) after brightness adjustment, mapping the gray value of the text image to a gray range (0-255) displayed by a display, namely:
wherein r ismax、rminThe maximum gray value and the minimum gray value of the text image are respectively, and f is the gray value of the mapped text image.
The step b) comprises the following steps:
b1) defining the maximum and minimum gray values in the text image as rmaxAnd rminThe initial value of the global threshold is
The initial value of the iteration times k is 0;
b2) according to the threshold value TkDividing the text image into two parts of target character and background, respectively calculating the pixel number of target character and backgroundAnd
and its gray level average value
And
then
b3) Solving a new threshold value:
b4) if T)k+1＝TkOr k is more than 100, ending; otherwise k is k +1, go to b 2);
b5) randomly selecting 100 points in the text image, and designating the number of times for calculating the stroke width v as an initial value of 0;
b6) setting a length threshold value length (hT) according to the size of a font in a text image, wherein the number n of the widths of effective character strokes is 0;
b7) if the pixel value f (x, y) < T of the point of interest (x, y)k+1Then, the black pixel points are extended from the point along the horizontal and vertical directions until leaving the target area, and the lengths Hl and Vl of the black pixel points in the horizontal and vertical directions are obtained through statistics, wherein v is v + 1.
b8) If H1 < length hT or Vl < length hT, taking the smaller length value as the width (n) of the character stroke of the point, wherein n is n + 1; otherwise, abandoning the point;
b9) if v > -100, exit; otherwise take a point down, go to b 7);
b10) and sequencing the widths of all effective target character strokes, and then taking the value of the widths as the width d of the strokes, so as to select the size w of the neighborhood calculation template, wherein the w is 2d + 1.
The step d) comprises the following steps:
d1) comparing the gray value f (x, y) of the investigation point (x, y) with a global threshold value Tk+1If the gray value f (x, y) of the point under consideration (x, y) is less than or equal to Tk+1Go to d 2); otherwise, the gray value g (x, y) of the target pixel point is 255, the next point is continuously scanned, and the process goes to d 1;
d2) finding the average gray level avg (x, y) in the w x w template taking the investigation point as the center;
d3) and d, comparing the gray value f (x, y) of the inspected point with the average gray value avg (x, y) obtained in the step d2), if the gray value f (x, y) of the inspected point is larger than the average gray value avg (x, y), the gray value g (x, y) of the target pixel point is 255, otherwise, the gray value g (x, y) of the target pixel point is 0, continuing to scan the next point, and turning to d 1).
The invention has the beneficial effects that: according to the invention, through global brightness adjustment of the text image, the gray contrast of the target character and the background in the text image can be improved, a text image with better quality is obtained, and preparation is made for subsequent binarization processing; in addition, the text image is amplified through a bicubic interpolation algorithm, so that the gap between character strokes can be increased, and the size of a calculation template in the following field can be conveniently determined; the size of the neighborhood calculation template is selected in a self-adaptive manner, so that the application range of the method is widened; in addition, the method of combining the global threshold value and the local threshold value is adopted to carry out point-by-point binarization processing on each character block, so that character strokes can be effectively segmented from the background, the phenomena of pen breaking, artifact and the like are avoided, the connectivity of the strokes is kept, the condition of calculating the local threshold value for pixel points in a text image is reduced, and the operation speed is greatly improved.
Drawings
FIG. 1 is a block diagram of a method for binarization processing of an adaptive text image based on pixel neighborhood characteristics according to the present invention;
FIG. 2 is a flow chart of a method for adaptively selecting a neighborhood calculation template according to the present invention;
FIG. 3 is a flow chart of the point-by-point binarization processing of the invention combining global and local thresholds;
FIG. 4 is a schematic diagram of a text image to be binarized according to the present invention;
FIG. 5 is a schematic diagram of a text image after global brightness adjustment according to the present invention;
FIG. 6 is a schematic diagram of a text image to be magnified according to the present invention;
fig. 7 is a schematic diagram of a text image after the point-by-point binarization processing combining global and local thresholds is performed.
Detailed Description
The invention is further elucidated with reference to the drawing.
As shown in fig. 1, the binarization processing method of the adaptive text image based on the pixel neighborhood characteristics of the present invention includes: 1) carrying out global brightness adjustment on the text image, and improving the gray scale contrast of the target character and the background in the text image; 2) carrying out zooming processing on the text image by adopting a bicubic interpolation algorithm; 3) adaptively selecting the size of a neighborhood calculation template; 4) dividing the text information of the text image into text blocks according to the size of the determined domain calculation template; 5) and performing point-by-point binarization processing on each character block by adopting a method of combining global and local threshold values.
The following steps are described in detail:
1) and (3) performing global brightness adjustment on the text image: the method comprises the following steps of adjusting the overall brightness of an original text image according to the overall brightness level of the original text image, increasing the proportion of a darker part in the image, and improving the gray scale contrast of target characters and a background in the text image, and comprises the following specific steps:
11) defining the pixel value of any point of the text image as the brightness value Y of the point, firstly carrying out normalization processing on the brightness value Y, wherein I represents the brightness value of the point after normalization,
I＝Y/255；
12) calculating the average value I of the gray scale of the whole text imageal，
Wherein rho epsilon I represents that rho points are in a definition domain of the text image, and N represents the number of pixel values in the text image;
13) a global luminance dynamic range compression degree coefficient gamma is defined,
wherein,
is the average value of the overall gray scale of the image IalThe increasing function of (1) represents the bending degree of the global dynamic range compression curve, the lower the average value is, the larger the bending degree of the curve is, the larger the stretching degree of the darker part is, but when the gray average value is larger, the gamma is taken as 1, and the image is not subjected to the whole dynamic range compression;
14) global brightness adjustment is carried out on the text image by utilizing the global brightness dynamic range compression degree coefficient gamma,
I′＝Iγ；
15) after brightness adjustment, mapping the gray value of the text image to a gray range (0-255) displayed by a display, namely:
wherein r ismax、rminThe maximum gray value and the minimum gray value of the text image are respectively, and f is the gray value of the mapped text image.
2) And (4) carrying out scaling processing on the text image by adopting a bicubic interpolation algorithm, and increasing gaps among character strokes.
3) And self-adaptively selecting the size of the neighborhood calculation template, and counting the stroke width d of the target character so as to determine the size w of the neighborhood calculation template.
Firstly, a global iterative algorithm is adopted to calculate a global threshold value, and the method firstly selects an initial threshold value
The initial value of the iteration number k is set to 0, wherein rmax、rminRespectively the maximum and minimum gray values in the image. According to the threshold value TkSegmenting an image into a target and a background (less than T)kAnd is not less than Tk) Two parts, respectively calculating the pixel numbers of the target part and the background part
And
and its gray level average value
And
the average expected value of the target and background is then used as a new threshold, and so on. When the threshold value is no longer changed, i.e. Tk+1＝TkOr when k is more than 100, stopping iteration, and generally reaching a stable state after several iterations.
Then, the width of the template is determined, 100 points are randomly selected in the text image, and the global threshold T is used for determining the width of the templatek+1Judging each point, and if the point is in the background area, not considering the point again; if the point is in the target area (black pixel), the point respectively extends along the horizontal direction and the vertical direction until the point leaves the target area, and the length of the black pixel in the horizontal direction and the vertical direction is obtained through statistics. Discarding the point when the length in both the vertical and horizontal directions exceeds a length threshold (set according to font size in the text image), otherwise taking the smaller of the twoThe small length value is used as the width of the character stroke of the point; the widths of all valid character strokes are sorted and then the value is taken as the width d of the stroke, thereby determining the width w of the template (w-2 d + 1).
As shown in fig. 2, the specific process is as follows:
31) defining the maximum and minimum gray values in the text image as rmaxAnd rminThe initial value of the global threshold is
The initial value of the iteration times k is 0;
32) according to the threshold value TkDividing the text image into two parts of target character and background, respectively calculating the pixel number of target character and background
And
and its gray level average valueAnd
then
33) Solving a new threshold value:
34) if T)k+1＝TkOr k is more than 100, ending; otherwise k is k +1, go to b 2);
35) randomly selecting 100 points in the text image, and designating the number of times for calculating the stroke width v as an initial value of 0;
36) setting a length threshold value length (hT) according to the size of a font in a text image, wherein the number n of the widths of effective character strokes is 0;
37) if the pixel value f (x, y) < T of the point of interest (x, y)k+1Then, the black pixel points are extended from the point along the horizontal and vertical directions until leaving the target area, and the lengths Hl and Vl of the black pixel points in the horizontal and vertical directions are obtained through statistics, wherein v is v + 1.
38) If Hl < length hT or Vl < length hT, taking the smaller length value as the width (n) of the character stroke of the point, wherein n is n + 1; otherwise, abandoning the point;
39) if v > -100, exit; otherwise take a point down, go to b 7);
310) and sequencing the widths of all effective target character strokes, and then taking the value of the widths as the width d of the strokes, so as to select the size w of the neighborhood calculation template, wherein the w is 2d + 1.
4) And dividing the character information of the text image into character blocks according to the size of the determined domain calculation template.
5) And performing point-by-point binarization processing on each character block by adopting a method of combining global and local threshold values. If the gray value f (x, y) of the point under consideration (x, y) is greater than Tk+1If yes, the gray value g (x, y) of the target pixel point is 255; otherwise, the average gray level avg (x, y) in the w x w template with the investigation point as the center is calculated, if the gray level f (x, y) of the investigation point is larger than the average gray level avg (x, y), the value of the gray level g (x, y) of the target pixel point is 255, otherwise, the gray level g (x, y) of the target pixel point is 0.
As shown in fig. 3, the specific process is as follows:
51) comparing the gray value f (x, y) of the investigation point (x, y) with a global threshold value Tk+1If the gray value f (x, y) of the point under investigation (x, y) is less than or equal toAt Tk+1Go to 52); otherwise, the gray value g (x, y) of the target pixel point is 255, the next point is continuously scanned, and 51 is turned to;
52) finding the average gray level avg (x, y) in the w x w template taking the investigation point as the center;
53) comparing the gray value f (x, y) of the inspected point with the average gray avg (x, y) obtained in the step 52), if the gray value f (x, y) of the inspected point is larger than the average gray avg (x, y), the gray value g (x, y) of the target pixel point is 255, otherwise, the gray value g (x, y) of the target pixel point is 0, continuing to scan the next point, and going to 51).
By adopting the method, the binary character image with higher quality can be obtained by traversing each pixel point of the original image.
The specific embodiment is as follows:
for the address partial image of the collected second generation resident identification document with the poor quality and the size of 377 x 164 pixels, as shown in fig. 4, firstly, the global brightness is adjusted in a self-adaptive manner to obtain an image with improved gray scale contrast of the target and the background, as shown in fig. 5; then, amplifying the image after brightness adjustment by adopting a bicubic interpolation algorithm according to a certain scale factor, thereby increasing the gap between character strokes, as shown in fig. 6; then counting the stroke width d of the target character for the graph 6, thereby determining the size w of the neighborhood calculation template; dividing the character information into character blocks according to the size of the template acquired by self-adaption; finally, point-by-point binarization is carried out on the image 6 in a text block by adopting a method of combining global and local threshold values, so that the target character can be clearly extracted from the background, noise is filtered, and a high-quality binarized text image is obtained, as shown in fig. 7, table 1 shows that OCR results of the image are compared with those of the image binarized by other binarizing algorithms.
TABLE 1
The above description is only a preferred embodiment of the present invention, the present invention is not limited to the above embodiment, and there may be some slight structural changes in the implementation, and if there are various changes or modifications to the present invention without departing from the spirit and scope of the present invention, and within the claims and equivalent technical scope of the present invention, the present invention is also intended to include those changes and modifications.
Claims (5)
1. A binarization processing method of a self-adaptive character image based on pixel neighborhood characteristics is characterized by comprising the following steps:
a) carrying out global brightness adjustment on the text image, and improving the gray scale contrast of the target character and the background in the text image;
b) adaptively selecting the size of a neighborhood calculation template;
c) dividing the text information of the text image into text blocks according to the size of the selected domain calculation template;
d) and performing point-by-point binarization processing on each character block by adopting a method of combining global and local threshold values.
2. The binarization processing method for the adaptive character image based on the pixel neighborhood characteristics as claimed in claim 1, wherein between the step a) and the step b), further comprising:
and ab) carrying out zooming processing on the text image by adopting a bicubic interpolation algorithm.
3. The binarization processing method for the adaptive character image based on the pixel neighborhood characteristics as claimed in claim 1, wherein the step a) comprises:
a1) defining the pixel value of any point of the text image as the brightness value Y of the point, normalizing the brightness value Y, I represents the brightness value of the point after normalization,
I＝Y/255；
a2) calculating the average value of the gray scale of the whole imageal，
Wherein rho epsilon I represents that rho points are in a definition domain of the text image, and N represents the number of pixel values in the text image;
a3) a global luminance dynamic range compression degree coefficient gamma is defined,
a4) global brightness adjustment is carried out on the text image by utilizing the global brightness dynamic range compression degree coefficient gamma,
I′＝Iγ；
a5) after brightness adjustment, mapping the gray value of the text image to a gray range (0-255) displayed by a display, namely:
wherein r ismax、rminThe maximum gray value and the minimum gray value of the text image are respectively, and f is the gray value of the mapped text image.
4. The binarization processing method for the adaptive character image based on the pixel neighborhood characteristics as claimed in claim 1, wherein the step b) comprises:
b1) defining the maximum and minimum gray values in the text image as rmaxAnd rminThe initial value of the global threshold is
The initial value of the iteration times k is 0;
b2) according to the threshold value TkDividing the text image into two parts of target character and background, respectively calculating the pixel number of target character and background
Andand its gray level average value
Andthen
b3) Solving a new threshold value:
b4) if T)k+1＝TkOr k is more than 100, ending; otherwise k is k +1, go to b 2);
b5) randomly selecting 100 points in the text image, and designating the number of times for calculating the stroke width v as an initial value of 0;
b6) setting a length threshold value length (hT) according to the size of a font in a text image, wherein the number n of the widths of effective character strokes is 0;
b7) if the pixel value f (x, y) < T of the point of interest (x, y)k+1Then, the black pixel points are extended from the point along the horizontal and vertical directions until leaving the target area, and the lengths Hl and Vl of the black pixel points in the horizontal and vertical directions are obtained through statistics, wherein v is v + 1.
b8) If H1 < length hT or Vl < length hT, taking the smaller length value as the width (n) of the character stroke of the point, wherein n is n + 1; otherwise, abandoning the point;
b9) if v > -100, exit; otherwise take a point down, go to b 7);
b10) and sequencing the widths of all effective target character strokes, and then taking the value of the widths as the width d of the strokes, so as to select the size w of the neighborhood calculation template, wherein the w is 2d + 1.
5. The binarization processing method for the adaptive character image based on the pixel neighborhood characteristics as claimed in claim 4, wherein the step d) comprises:
d1) comparing the gray value f (x, y) of the investigation point (x, y) with a global threshold value Tk+1If the gray value f (x, y) of the point under consideration (x, y) is less than or equal to Tk+1Go to d 2); otherwise, the gray value g (x, y) of the target pixel point is 255, the next point is continuously scanned, and the process goes to d 1;
d2) finding the average gray level avg (x, y) in the w x w template taking the investigation point as the center;
d3) and d, comparing the gray value f (x, y) of the inspected point with the average gray value avg (x, y) obtained in the step d2), if the gray value f (x, y) of the inspected point is larger than the average gray value avg (x, y), the gray value g (x, y) of the target pixel point is 255, otherwise, the gray value g (x, y) of the target pixel point is 0, continuing to scan the next point, and turning to d 1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102641444A CN102289668A (en) | 2011-09-07 | 2011-09-07 | Binaryzation processing method of self-adaption word image based on pixel neighborhood feature |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102641444A CN102289668A (en) | 2011-09-07 | 2011-09-07 | Binaryzation processing method of self-adaption word image based on pixel neighborhood feature |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102289668A true CN102289668A (en) | 2011-12-21 |
Family
ID=45336074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011102641444A Pending CN102289668A (en) | 2011-09-07 | 2011-09-07 | Binaryzation processing method of self-adaption word image based on pixel neighborhood feature |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102289668A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103823863A (en) * | 2014-02-24 | 2014-05-28 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN104637047A (en) * | 2013-11-13 | 2015-05-20 | 北京慧眼智行科技有限公司 | Image processing method and device |
| CN105279727A (en) * | 2014-07-08 | 2016-01-27 | 腾讯科技（深圳）有限公司 | Image processing method and apparatus |
| CN105844608A (en) * | 2015-01-16 | 2016-08-10 | 西门子医疗保健诊断公司 | Urinary sediment image segmentation method and urinary sediment image segmentation device |
| CN106651984A (en) * | 2016-01-21 | 2017-05-10 | 上海联影医疗科技有限公司 | Computer tomography artifact correction method and device |
| CN106778732A (en) * | 2017-01-16 | 2017-05-31 | 哈尔滨理工大学 | Text information feature extraction and recognition method based on Gabor filter |
| CN107895355A (en) * | 2017-11-30 | 2018-04-10 | 天津天地基业科技有限公司 | A kind of mobile detection and picture contrast system for adaptive enhancement and its method |
| CN105335953B (en) * | 2014-07-07 | 2018-04-10 | 富士通株式会社 | Extract the apparatus and method of the background luminance figure of image, go shade apparatus and method |
| CN108288064A (en) * | 2017-01-09 | 2018-07-17 | 北京京东尚科信息技术有限公司 | Method and apparatus for generating picture |
| CN108628858A (en) * | 2018-04-20 | 2018-10-09 | 广东科学技术职业学院 | The operating method and system of textual scan identification translation on line based on mobile terminal |
| CN109344297A (en) * | 2018-09-18 | 2019-02-15 | 北京工业大学 | The method of CIP data is obtained in a kind of shared book system offline |
| CN110263301A (en) * | 2019-06-27 | 2019-09-20 | 北京百度网讯科技有限公司 | Method and apparatus for determining the color of text |
| CN110288618A (en) * | 2019-04-24 | 2019-09-27 | 广东工业大学 | A kind of Segmentation of Multi-target method of uneven illumination image |
| CN110502950A (en) * | 2019-08-09 | 2019-11-26 | 广东技术师范大学 | A kind of quick self-adapted binarization method of QR code of uneven illumination |
| CN110596746A (en) * | 2019-10-17 | 2019-12-20 | 中国测试技术研究院辐射研究所 | Method for using an automatic test/calibration/verification device for a dose equivalent instrument |
| CN111986222A (en) * | 2020-08-21 | 2020-11-24 | 国网宁夏电力有限公司营销服务中心(国网宁夏电力有限公司计量中心) | Intelligent electric meter chip image binarization processing method based on self-adaptive mixed threshold value |
| CN112861794A (en) * | 2021-03-11 | 2021-05-28 | 浙江康旭科技有限公司 | Universal detection algorithm for optical printing texts and scene texts |
| CN116485924A (en) * | 2023-03-20 | 2023-07-25 | 西安电子科技大学 | Binarization method of CT section image of optical fiber coil containing artifact |
| WO2024027583A1 (en) * | 2022-08-03 | 2024-02-08 | 维沃移动通信有限公司 | Image processing method and apparatus, and electronic device and readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040052427A1 (en) * | 2002-07-01 | 2004-03-18 | Xerox Corporation | Foreground erosion method and system for Multiple Raster Content (MRC) representation of documents |
| CN1941838A (en) * | 2005-09-29 | 2007-04-04 | 株式会社理光 | File and picture binary coding method |
-
2011
- 2011-09-07 CN CN2011102641444A patent/CN102289668A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040052427A1 (en) * | 2002-07-01 | 2004-03-18 | Xerox Corporation | Foreground erosion method and system for Multiple Raster Content (MRC) representation of documents |
| CN1941838A (en) * | 2005-09-29 | 2007-04-04 | 株式会社理光 | File and picture binary coding method |
Non-Patent Citations (1)
| Title |
|---|
| 庄军 等: "一种有效的文本图像二值化方法", 《微计算机信息》, vol. 21, no. 8, 31 August 2005 (2005-08-31) * |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104637047A (en) * | 2013-11-13 | 2015-05-20 | 北京慧眼智行科技有限公司 | Image processing method and device |
| CN103823863B (en) * | 2014-02-24 | 2017-07-25 | 联想(北京)有限公司 | A kind of information processing method and electronic equipment |
| CN103823863A (en) * | 2014-02-24 | 2014-05-28 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN105335953B (en) * | 2014-07-07 | 2018-04-10 | 富士通株式会社 | Extract the apparatus and method of the background luminance figure of image, go shade apparatus and method |
| CN105279727B (en) * | 2014-07-08 | 2019-08-06 | 腾讯科技（深圳）有限公司 | Image processing method and device |
| CN105279727A (en) * | 2014-07-08 | 2016-01-27 | 腾讯科技（深圳）有限公司 | Image processing method and apparatus |
| CN105844608A (en) * | 2015-01-16 | 2016-08-10 | 西门子医疗保健诊断公司 | Urinary sediment image segmentation method and urinary sediment image segmentation device |
| CN106651984A (en) * | 2016-01-21 | 2017-05-10 | 上海联影医疗科技有限公司 | Computer tomography artifact correction method and device |
| CN108288064A (en) * | 2017-01-09 | 2018-07-17 | 北京京东尚科信息技术有限公司 | Method and apparatus for generating picture |
| CN106778732A (en) * | 2017-01-16 | 2017-05-31 | 哈尔滨理工大学 | Text information feature extraction and recognition method based on Gabor filter |
| CN107895355B (en) * | 2017-11-30 | 2021-08-20 | 天津天地基业科技有限公司 | Motion detection and image contrast self-adaptive enhancement system and method |
| CN107895355A (en) * | 2017-11-30 | 2018-04-10 | 天津天地基业科技有限公司 | A kind of mobile detection and picture contrast system for adaptive enhancement and its method |
| CN108628858A (en) * | 2018-04-20 | 2018-10-09 | 广东科学技术职业学院 | The operating method and system of textual scan identification translation on line based on mobile terminal |
| CN109344297A (en) * | 2018-09-18 | 2019-02-15 | 北京工业大学 | The method of CIP data is obtained in a kind of shared book system offline |
| CN110288618B (en) * | 2019-04-24 | 2022-09-23 | 广东工业大学 | Multi-target segmentation method for uneven-illumination image |
| CN110288618A (en) * | 2019-04-24 | 2019-09-27 | 广东工业大学 | A kind of Segmentation of Multi-target method of uneven illumination image |
| CN110263301A (en) * | 2019-06-27 | 2019-09-20 | 北京百度网讯科技有限公司 | Method and apparatus for determining the color of text |
| CN110263301B (en) * | 2019-06-27 | 2023-12-05 | 北京百度网讯科技有限公司 | Method and device for determining color of text |
| CN110502950A (en) * | 2019-08-09 | 2019-11-26 | 广东技术师范大学 | A kind of quick self-adapted binarization method of QR code of uneven illumination |
| CN110596746A (en) * | 2019-10-17 | 2019-12-20 | 中国测试技术研究院辐射研究所 | Method for using an automatic test/calibration/verification device for a dose equivalent instrument |
| CN110596746B (en) * | 2019-10-17 | 2024-03-01 | 中国测试技术研究院辐射研究所 | Method for automatic testing/calibrating device using dose equivalent instrument |
| CN111986222A (en) * | 2020-08-21 | 2020-11-24 | 国网宁夏电力有限公司营销服务中心(国网宁夏电力有限公司计量中心) | Intelligent electric meter chip image binarization processing method based on self-adaptive mixed threshold value |
| CN112861794A (en) * | 2021-03-11 | 2021-05-28 | 浙江康旭科技有限公司 | Universal detection algorithm for optical printing texts and scene texts |
| WO2024027583A1 (en) * | 2022-08-03 | 2024-02-08 | 维沃移动通信有限公司 | Image processing method and apparatus, and electronic device and readable storage medium |
| CN116485924A (en) * | 2023-03-20 | 2023-07-25 | 西安电子科技大学 | Binarization method of CT section image of optical fiber coil containing artifact |
| CN116485924B (en) * | 2023-03-20 | 2023-09-29 | 西安电子科技大学 | Binarization method of CT section image of optical fiber coil containing artifact |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102289668A (en) | Binaryzation processing method of self-adaption word image based on pixel neighborhood feature | |
| CN107527333B (en) | Quick image enhancement method based on gamma transformation | |
| CN109658424B (en) | Improved robust two-dimensional OTSU threshold image segmentation method | |
| US9251614B1 (en) | Background removal for document images | |
| KR101795823B1 (en) | Text enhancement of a textual image undergoing optical character recognition | |
| CN105513035B (en) | The detection method and its system of human body concealment article in a kind of passive millimeter wave image | |
| CN105374015A (en) | Binary method for low-quality document image based on local contract and estimation of stroke width | |
| CN111145105B (en) | Image rapid defogging method and device, terminal and storage medium | |
| CN109241973B (en) | Full-automatic soft segmentation method for characters under texture background | |
| CN108830857B (en) | Self-adaptive Chinese character copy label image binarization segmentation method | |
| CN113592776B (en) | Image processing method and device, electronic equipment and storage medium | |
| CN109272461A (en) | Infrared image enhancing method based on median filtering and color histogram | |
| CN113781406B (en) | Scratch detection method and device for electronic component and computer equipment | |
| CN104463814A (en) | Image enhancement method based on local texture directionality | |
| CN112381826B (en) | Binarization method of edge defect image | |
| CN104361335B (en) | A kind of processing method that black surround is automatically removed based on scan image | |
| CN112288726A (en) | Method for detecting foreign matters on belt surface of underground belt conveyor | |
| CN101599172A (en) | The illumination compensation splitting method of the text image of inhomogeneous illumination | |
| JP2008210387A (en) | Noise elimination device and noise elimination program for improving binarization performance of document image | |
| CN108205678B (en) | Nameplate character recognition processing method containing bright spot interference | |
| CN104268845A (en) | Self-adaptive double local reinforcement method of extreme-value temperature difference short wave infrared image | |
| CN104616259A (en) | Non-local mean image de-noising method with noise intensity self-adaptation function | |
| Saini | Document image binarization techniques, developments and related issues: a review | |
| CN110930358B (en) | Solar panel image processing method based on self-adaptive algorithm | |
| Tabatabaei et al. | A novel method for binarization of badly illuminated document images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: GUANGZHOU KANSIG ELECTRONICS TECHNOLOGY INC. Free format text: FORMER OWNER: TAN HONGZHOU Effective date: 20121227 Free format text: FORMER OWNER: ZHU XIONGYONG YANG JIN Effective date: 20121227 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 510006 GUANGZHOU, GUANGDONG PROVINCE TO: 510663 GUANGZHOU, GUANGDONG PROVINCE |
|
| TA01 | Transfer of patent application right |
Effective date of registration: 20121227 Address after: 510663 Guangdong city of Guangzhou province high tech Industrial Development Zone of Guangzhou Science City on Road No. 80, Guangzhou technology innovation base E District second floor 202 unit Applicant after: Guangzhou Kansig Electronics Technology Inc. Address before: 510006 laboratory room of information technology and technology, Zhongshan University, Panyu District University Town, Guangzhou, Guangdong 311, China Applicant before: Tan Hongzhou Applicant before: Zhu Xiongyong Applicant before: Yang Jin |
|
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20111221 |