CN103593329A - Text image rearrangement method and system - Google Patents
Text image rearrangement method and system Download PDFInfo
- Publication number
- CN103593329A CN103593329A CN201210295562.4A CN201210295562A CN103593329A CN 103593329 A CN103593329 A CN 103593329A CN 201210295562 A CN201210295562 A CN 201210295562A CN 103593329 A CN103593329 A CN 103593329A
- Authority
- CN
- China
- Prior art keywords
- text
- pixel
- line
- column
- word picture
- 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.)
- Granted
Links
Images
Abstract
The invention relates to a text image rearrangement method and system. The text image rearrangement method includes dividing pixels in a text image to be rearranged into background pixels and text pixels through image binarization; determining text lines of the text image to be rearranged according to the text pixels; determining text columns of the text lines according to the text pixels and the text lines of the text image to be rearranged; determining a paragraph in which the text lines of the text image to be rearranged located according to spacing and line height between adjacent text lines; rearranging the text columns in the text image to be rearranged according to the preset size of rearranged text image, the text columns and a paragraph in which the text lines are located, and obtaining the rearranged text image. The invention further relates to a text image rearrangement system. According to the text image rearrangement method and system, text images can be rearranged, and the rearranged text images are adaptable to being browed on a mobile phone.
Description
Technical field
The present invention relates to picture processing field, particularly relate to a kind of word picture rearrangement method and system.
Background technology
Along with social development, increasing people likes using surfing Internet with cell phone browsing page and exchange and interdynamic.But because some picture design in website is only to have considered computer user, be therefore only applicable to use computer screen to watch.Especially some novel website, in order to prevent that other websites from copying novel word, is all converted to picture format (being word picture) by novel word.When user uses mobile phone to watch, if do not amplified corresponding word picture, can cannot see the word in word picture clearly because of too small causing of mobile phone screen.As directly corresponding word picture amplified to processing, because the size of word picture is much larger than mobile phone screen size, can cause user need to drag continually word picture in order to browse word picture, affect user's word picture browsing.
Therefore, be necessary to provide a kind of word picture rearrangement method and system, to solve the existing problem of prior art.
Summary of the invention
The object of the present invention is to provide a kind of word picture rearrangement method and system, solved the technical matters of mobile phone user word picture inconvenience.
For addressing the above problem, technical scheme provided by the invention is as follows:
The present invention relates to a kind of word picture rearrangement method, it comprises:
By image binaryzation, each pixel in word picture to be reset is divided into background pixel and text pixel;
According to described text pixel, the line of text of word picture to be reset described in determining;
According to described text pixel and described in the line of text of word picture to be reset, determine the text column of described line of text;
High according to the spacing of adjacent described line of text and row, the paragraph at the line of text place of word picture to be reset described in determining; And
According to the paragraph at the size of the word picture after default rearrangement, described text column and described text column place, the text column in word picture described to be reset is reset to processing, obtain the word picture after described rearrangement.
In word picture rearrangement method of the present invention, the described step that all pixels in word picture to be reset is divided into background pixel and text pixel by image binaryzation comprises:
The pixel gradient value of each pixel described in calculating in word picture to be reset,
The mean value of the pixel gradient value of each pixel described in setting in word picture to be reset is pixel gradient mean value, the described pixel that described pixel gradient value is more than or equal to described pixel gradient mean value is set as described text pixel, and the described pixel that described pixel gradient value is less than to described pixel gradient mean value is set as described background pixel.
In word picture rearrangement method of the present invention, the described step that all pixels in word picture to be reset is divided into background pixel and text pixel by image binaryzation comprises:
The grey scale pixel value of choosing variance maximum in each pixel grey scale variance is gray scale segmentation threshold;
According to described gray scale segmentation threshold, all pixels in word picture described to be reset are divided into and are less than or equal to the pixel of described gray scale segmentation threshold and pixel two classes that gray-scale value is greater than described gray scale segmentation threshold, the class pixel that number is many is set as described text pixel, and the class pixel that number is few is set as described background pixel.
In word picture rearrangement method of the present invention, before the described step that by image binaryzation, all pixels in word picture to be reset is divided into background pixel and text pixel, also comprise:
Word picture described to be reset is converted to gray scale picture.
In word picture rearrangement method of the present invention, described by image binaryzation, all pixels in word picture to be reset are divided into the step of background pixel and text pixel after, described according to described text pixel, described in determining, before the step of the line of text of word picture to be reset, also comprise:
As all neighbors of processed text pixel be as described in background pixel, change described processed text pixel into described background pixel.
In word picture rearrangement method of the present invention, described according to described text pixel, described in determining, the step of the line of text of word picture to be reset comprises:
By all pixels of word picture to be reset described in line scanning, as be scanned in capable pixel as described in the number of text pixel be greater than 0, will described in be scanned row and be set as text pixel column, otherwise will described in be scanned row to be set as background pixel capable;
Adjacent described text pixel column is merged to processing, obtain preliminary line of text;
High according to the row of described preliminary line of text, described preliminary line of text is merged to processing, obtain described line of text.
In word picture rearrangement method of the present invention, described according to described text pixel and described in the line of text of word picture to be reset, determine that the step of the text column of described line of text comprises:
By the line of text of word picture to be reset described in column scan, as be scanned in the pixel of row as described in the number of text pixel be greater than 0, by described, be scanned the text pixel column that row are set as described line of text, otherwise described being scanned is listed as to the background pixel row that are set as described line of text;
Adjacent described text pixel column is merged to processing, obtain the preliminary text column of described line of text;
According to the col width of the preliminary text column of described line of text, described preliminary text column is merged to processing, obtain the text column of described line of text.
In word picture rearrangement method of the present invention, described according to the paragraph at size, described text column and the described text column place of the word picture after resetting, text column in word picture described to be reset is reset to processing, and the step that obtains the word picture after described rearrangement comprises:
Size according to the word picture after default rearrangement, creates a background picture;
Successively all text column of each paragraph are reset on described background picture line by line, and corresponding paragraph marks is set, obtain the word picture after described rearrangement.
The invention still further relates to a kind of word picture system for rearranging, it comprises:
Divide module, for each pixel of word picture to be reset being divided into background pixel and text pixel by image binaryzation;
Line of text determination module, for according to described text pixel, the line of text of word picture to be reset described in determining;
Text column determination module, for according to described text pixel and described in the line of text of word picture to be reset, determine the text column of described line of text;
Text fragment determination module, for high according to the spacing of adjacent described line of text and row, the paragraph at the line of text place of word picture to be reset described in determining; And
Reordering module, for according to the paragraph at the size of the word picture after default rearrangement, described text column and described text column place, resets processing to the text column in word picture described to be reset, and obtains the word picture after described rearrangement.
In word picture system for rearranging of the present invention, described division module comprises:
Gradient division unit, for the pixel gradient value of each pixel of word picture to be reset described in calculating;
The mean value of the pixel gradient value of each pixel described in setting in word picture to be reset is pixel gradient mean value, the described pixel that described pixel gradient value is more than or equal to described pixel gradient mean value is set as described text pixel, and the described pixel that described pixel gradient value is less than to described pixel gradient mean value is set as described background pixel; And
Between class, poor division unit, is gray scale segmentation threshold for choosing the grey scale pixel value of each pixel grey scale variance variance maximum;
According to described gray scale segmentation threshold, all pixels in word picture described to be reset are divided into and are less than or equal to the pixel of described gray scale segmentation threshold and pixel two classes that gray-scale value is greater than described gray scale segmentation threshold, the class pixel that number is many is set as described text pixel, and the class pixel that number is few is set as described background pixel.
In word picture system for rearranging of the present invention, described word picture system for rearranging also comprises:
Picture modular converter, for being converted to gray scale picture by word picture described to be reset; And
Pixel denoising module, for described text pixel is carried out to denoising, as all neighbors of processed text pixel be as described in background pixel, change described processed text pixel into described background pixel.
In word picture system for rearranging of the present invention, described line of text determination module comprises:
Pixel column identifying unit, be used for by all pixels of word picture to be reset described in line scanning, as be scanned in capable pixel as described in the number of text pixel be greater than 0, will described in be scanned row and be set as text pixel column, otherwise will described in be scanned row to be set as background pixel capable;
Preliminary line of text generation unit, for adjacent described text pixel column is merged to processing, obtains preliminary line of text; And
Line of text generation unit, high for the row according to described preliminary line of text, described preliminary line of text is merged to processing, obtain described line of text.
In word picture system for rearranging of the present invention, described text column determination module comprises:
Pixel column identifying unit, be used for by the line of text of word picture to be reset described in column scan, as be scanned in the pixel of row as described in the number of text pixel be greater than 0, by described, be scanned the text pixel column that row are set as described line of text, otherwise described being scanned is listed as to the background pixel row that are set as described line of text;
Preliminary text column generation unit, for adjacent described text pixel column is merged to processing, obtains the preliminary text column of described line of text; And
Text column generation unit, for according to the col width of the preliminary text column of described line of text, merges processing to described preliminary text column, obtains the text column of described line of text.
In word picture system for rearranging of the present invention, described reordering module comprises:
Background creating unit, for according to the size of the word picture after default rearrangement, creates a background picture; And
Unit reset in word, for successively all text column of each paragraph being reset to described background picture line by line, and corresponding paragraph marks is set, and obtains the word picture after described rearrangement.
Compared to prior art, word picture rearrangement method of the present invention and system are reset processing to word picture, and reset word picture and mobile phone screen sizableness after processing, make the word picture of resetting after processing be adapted at browsing on mobile phone, simplify operation when cellphone subscriber reads word, solved the technical matters of mobile phone user word picture inconvenience.
For foregoing of the present invention can be become apparent, preferred embodiment cited below particularly, and coordinate appended graphicly, be described in detail below:
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the preferred embodiment of word picture rearrangement method of the present invention;
One of dot structure that Fig. 2 is judgement neighbor;
Two of the dot structure that Fig. 3 is judgement neighbor;
Fig. 4 is the particular flow sheet of the step 102 in process flow diagram shown in Fig. 1;
Fig. 5 A is for adopting the word picture rearrangement method of invention to merge the schematic diagram before processing to adjacent text pixel column;
Fig. 5 B is for adopting word picture rearrangement method of the present invention to merge the schematic diagram after processing to adjacent text pixel column;
Fig. 6 is the particular flow sheet of the step 103 in process flow diagram shown in Fig. 1;
Fig. 7 A is for adopting the word picture rearrangement method of invention to merge the schematic diagram before processing to adjacent text pixel column;
Fig. 7 B is for adopting word picture rearrangement method of the present invention to merge the schematic diagram after processing to adjacent text pixel column;
Fig. 8 is the particular flow sheet of the step 104 in process flow diagram shown in Fig. 1;
Fig. 9 is the particular flow sheet of the step 105 in process flow diagram shown in Fig. 1;
Figure 10 is the generation schematic diagram of the word picture after the rearrangement of step 105 of word picture rearrangement method of the present invention;
Figure 11 is the structural representation of the preferred embodiment of word picture system for rearranging of the present invention;
Wherein, description of reference numerals is as follows:
111, picture modular converter;
112, divide module;
113, pixel denoising module;
114, line of text determination module;
115, text column determination module;
116, text fragment determination module;
117, reordering module.
Embodiment
The explanation of following embodiment is graphic with reference to what add, can be in order to the specific embodiment of implementing in order to illustration the present invention.The direction term that the present invention mentions, such as " on ", D score, 'fornt', 'back', " left side ", " right side ", " interior ", " outward ", " side " etc., be only the direction with reference to annexed drawings.Therefore, the direction term of use is in order to illustrate and to understand the present invention, but not in order to limit the present invention.
In the drawings, the unit of structural similarity is to represent with same numeral.
Please refer to Fig. 1, Fig. 1 is the process flow diagram of the preferred embodiment of word picture rearrangement method of the present invention.This word picture rearrangement method starts from:
Step 101, is divided into background pixel and text pixel by image binaryzation by all pixels in word picture to be reset;
Step 105, according to the paragraph at the size of the word picture after default rearrangement, text column and text column place, the text column for the treatment of in the word picture of rearrangement is reset processing, the word picture after being reset.
This word picture rearrangement method ends at step 105.
Below in conjunction with other accompanying drawings, describe the idiographic flow of each step in word picture rearrangement method of the present invention in detail.
In step 101, first judge whether this word picture to be reset is colour picture, as this is converted to after gray scale picture wait the word picture of resetting for colour picture, then treat all pixels in the word picture of rearrangement and divide.
For form, be each pixel in the word picture to be reset of gray scale picture, preferably adopt the gradient of pixel grey scale to divide it, be specially:
By following formula, calculate the pixel gradient value of each pixel in word picture to be reset,
Wherein G (x, y) is the gray-scale value of the described pixel located of coordinate (x, y), and Gr (x, y) is the pixel gradient value of the described pixel located of coordinate (x, y);
The mean value of setting the pixel gradient value of all pixels in word picture to be reset is pixel gradient mean value, the pixel that pixel gradient value is more than or equal to pixel gradient mean value is set as text pixel, and the pixel that pixel gradient value is less than to pixel gradient mean value is set as background pixel.
So, by the gradient of pixel grey scale, all pixels in word picture to be reset are divided for background pixel and text pixel, but the reason such as may be due to the difference of the pixel grey scale of text pixel and background pixel when the gradient that adopts pixel grey scale is divided too little, causes pixel to divide mistake.At this moment as all text pixel detected and be interconnected, judgement adopts the gradient of pixel grey scale to carry out pixel to divide unsuccessfully, and the inter-class variance that carries out following employing pixel grey scale carries out the division of pixel.
The method that the inter-class variance of employing pixel grey scale carries out the division of pixel is specially,
Set a gray scale segmentation threshold, make each pixel grey scale variance of S(of following formula) obtain maximal value;
Wherein si is the number that gray-scale value is less than or equal to the pixel of gray scale segmentation threshold; Msi is the average gray value that gray-scale value is less than or equal to all pixels of gray scale segmentation threshold; Bi is the number that gray-scale value is greater than the pixel of gray scale segmentation threshold; Mbi is the average gray value that gray-scale value is greater than all pixels of gray scale segmentation threshold;
As si is less than bi, the pixel that gray-scale value is less than or equal to gray scale segmentation threshold is set as text pixel, and the pixel that gray-scale value is greater than to described gray scale segmentation threshold is set as background pixel;
As si is greater than bi, the pixel that gray-scale value is greater than to described gray scale segmentation threshold is set as text pixel, and the pixel that gray-scale value is less than or equal to gray scale segmentation threshold is set as background pixel.
All pixels in word picture to be reset are divided for background pixel and text pixel so the most at last, certainly also can adopt additive method to divide pixel here, the method that for example color space quantizes (is divided into two classes by all pixels of word picture to be reset by color, the class that number of pixels is many is made as background pixel, and the class that number of pixels is few is made as text pixel) etc.
As shown in Figures 2 and 3, one of dot structure that Fig. 2 is judgement neighbor, two of the dot structure that Fig. 3 is judgement neighbor.In order to remove the impact of picture noise, the text pixel after dividing is carried out to denoising, be specially:
Detect successively all text pixel, as all neighbors of detected text pixel are background pixel, change this detected text pixel into background pixel.Here for the neighbor judging, can be that four neighbors of upper and lower, left and right of detected text pixel are (if the x of Fig. 2 is detected text pixel, 0,1,2,3 is four neighbors), also can be eight neighbors (if the x in Fig. 3 is text pixel, 0,1,2,3,4,5,6,7 is eight neighbors) of upper and lower, left and right, upper left, upper right, bottom left, bottom right etc.The number that neighbor Chinese word pixel also can be set certainly is here less than at 1 o'clock, changes this detected text pixel into background pixel etc.Concrete neighbor setting and the judgement of denoising can need to arrange according to user.
Please refer to Fig. 4, Fig. 4 is the particular flow sheet of the step 102 in process flow diagram shown in Fig. 1.Step 102 comprises:
Step 1021, by all pixels of line scanning word picture to be reset, as the number that is scanned the text pixel in capable pixel is greater than 0, this is scanned to row and is set as text pixel column, otherwise will be scanned, capable to be set as background pixel capable.
In this step, set H
avefor the average row of all preliminary line of text is high, hr
minfor the high minimum scale coefficient of row, hr
maxfor the high maximum ratio coefficient of row, if the row higher primary school of preliminary line of text is in H
ave* hr
min, and the row of the preliminary line of text of preliminary line of text and adjacent lastrow after merging is high, and the row senior middle school of the preliminary line of text of preliminary line of text and adjacent next line after merging at least one be less than H
ave* hr
max, the high less adjacent preliminary line of text of the row by preliminary line of text and after merging merges, and repeats this step until merge without preliminary line of text, subsequently all preliminary line of text is set as to line of text.
Because a line of text may comprise a plurality of preliminary line of text, line of text may be separated by preliminary background row and forms a plurality of preliminary line of text.In this step, by detecting the row of preliminary line of text is high line by line, the preliminary line of text of one text row is merged.Here be provided with two decision conditions, (row higher primary school is in H to merge at once condition
ave* hr
min) and row merges cancellation condition, and (row is tall and big in H
ave* hr
max).
No enough row as high in the row of preliminary line of text merges condition, this preliminary line of text is not merged.
As highly in the row of preliminary line of text meet row merging condition, the high no enough row of row of the preliminary line of text after the preliminary line of text of this preliminary line of text and adjacent lastrow merges merges cancellation condition, the high satisfied row of row of the preliminary line of text after the preliminary line of text of this preliminary line of text and adjacent next line merges merges cancellation condition, the preliminary line of text of this preliminary line of text and adjacent lastrow is merged.
As highly in the row of preliminary line of text meet row merging condition, the high no enough row of row of the preliminary line of text after the preliminary line of text of this preliminary line of text and adjacent next line merges merges cancellation condition, the high satisfied row of row of the preliminary line of text after the preliminary line of text of this preliminary line of text and adjacent lastrow merges merges cancellation condition, the preliminary line of text of this preliminary line of text and adjacent next line is merged.
As highly in the row of preliminary line of text meet row merging condition, the high no enough row of row of the preliminary line of text after the preliminary line of text of this preliminary line of text and adjacent lastrow merges merges cancellation condition, the row height of the preliminary line of text after the preliminary line of text of this preliminary line of text and adjacent next line merges also no enough row merges cancellation condition, and at this moment the high less adjacent preliminary line of text of the row of the preliminary line of text by preliminary line of text and after merging merges.
After having detected all preliminary line of text, more again the preliminary line of text after all merging is detected, the preliminary line of text meeting after the merging of above-mentioned condition is merged to processing again.So cycle detection, merging, until the preliminary line of text of nothing need to be carried out union operation, are set as line of text by all preliminary line of text subsequently.
The high minimum scale coefficient h of the row r here
minwith the high maximum ratio coefficient h r of row
maxcan set according to user's concrete needs, in the present embodiment capable high minimum scale coefficient h r
minbe set as 0.4, the high maximum ratio coefficient h r of row
maxbe set as 1.2.
Please refer to Fig. 6, Fig. 6 is the particular flow sheet of the step 103 in process flow diagram shown in Fig. 1.Step 103 comprises:
Step 1032, detects all text column in each line of text successively, left right continuous text column in line of text is set as to a text pixel column groups of this article one's own profession, sets Wn
maxfor the maximum text pixel number of the text pixel column in text pixel column groups, in text pixel column groups, there is the text pixel number of the text pixel column of maximum text pixel number; Wr
tfor row denoising scale-up factor.As the number of the text pixel of the text pixel column in text pixel column groups is less than Wn
max* Wr
t, the background pixel that text pixel column is set as to this article one's own profession is listed as, and can further eliminate like this interference of picture noise to the judgement of character image Chinese word pixel.The row denoising scale-up factor Wr here
tcan set according to user's concrete needs the scale-up factor of row denoising in the present embodiment Wr
tbe set as 0.11.
Step 1034, according to the col width of the preliminary text column of line of text, merges processing to preliminary text column, obtains the text column of this article one's own profession.
In this step, set W
maxfor the maximum col width of the text column in line of text, the i.e. col width of the text column of the maximum col width in this article one's own profession, hr
minfor col width minimum scale coefficient, wr
maxfor col width maximum ratio coefficient, as the col width of preliminary text column is less than W
max* wr
min, and the col width of the preliminary text column of the preliminary text column of preliminary text column and adjacent the first from left row after merging, and in the col width of the preliminary text column of the preliminary text column of preliminary text column and adjacent right row after merging, at least one is less than W
max* wr
max, the adjacent preliminary text column that col width of the preliminary text column by preliminary text column and after merging is less merges, and repeats this step until merge without preliminary text column, subsequently all preliminary text column is set as to the text column of this article one's own profession.
Because a text column may comprise a plurality of preliminary text column, text column may be separated by preliminary background row and forms a plurality of preliminary text column.In this step, by the col width to preliminary text column by column, detect, the preliminary text column of one text row is merged.Here be provided with two decision conditions, (col width is less than W to be listed as merging condition
max* wr
min) and row merge cancellation condition, and (col width is greater than W
max* wr
max).
As the col width of preliminary text column does not meet row merging condition, this preliminary text column is not merged.
As the col width of preliminary text column meets row merging condition, the col width of the preliminary text column after the preliminary text column of this preliminary text column and adjacent the first from left row merges does not meet row and merges cancellation condition, the col width of the preliminary text column after the preliminary text column of this preliminary text column and adjacent right row merges meets row and merges cancellation condition, the preliminary text column of this preliminary text column and adjacent the first from left row is merged.
As the col width of preliminary text column meets row merging condition, the col width of the preliminary text column after the preliminary text column of this preliminary text column and adjacent right row merges does not meet row and merges cancellation condition, the col width of the preliminary text column after the preliminary text column of this preliminary text column and adjacent the first from left row merges meets row and merges cancellation condition, the preliminary text column of this preliminary text column and adjacent right row is merged.
As the col width of preliminary text column meets row merging condition, the col width of the preliminary text column after the preliminary text column of this preliminary text column and adjacent the first from left row merges does not meet row and merges cancellation condition, the col width of the preliminary text column after the preliminary text column of this preliminary text column and adjacent right row merges does not meet row yet and merges cancellation condition, and at this moment the less adjacent preliminary text column of col width of the preliminary text column by preliminary text column and after merging merges.
After having detected all preliminary text column of a line of text, more again the preliminary text column after all merging is detected, the preliminary text column meeting after the merging of above-mentioned condition is merged to processing again.So cycle detection, merging, until the preliminary text column of nothing need to be carried out union operation, are set as all preliminary text column the text column of this article one's own profession subsequently.
The col width minimum scale coefficient wr here
minwith col width maximum ratio coefficient wr
maxcan set according to user's concrete needs, in the present embodiment col width minimum scale coefficient wr
minbe set as 0.4, col width maximum ratio coefficient wr
maxbe set as 1.2.
Subsequently next line of text is carried out the judgement of text column, until the text column in all line of text in this word picture to be reset is judged, finish.
Please refer to Fig. 8, Fig. 8 is the particular flow sheet of the step 104 in process flow diagram shown in Fig. 1.Step 104 comprises:
First calculate the spacing mean value (row that is the background row between adjacent line of text is high) of all line of text of word picture to be reset.The adjacent line of text of spacing mean value that spacing is more than or equal to all line of text of word picture to be reset is set as the adjacent line of text of large spacing, and the adjacent line of text of spacing mean value that spacing is less than to all line of text of word picture to be reset is set as the adjacent line of text of little spacing.Then calculate the spacing mean value Mb of the adjacent line of text of all large spacing
ave, and the spacing mean value Ms of the adjacent line of text of all little spacing
ave.As the space D of adjacent two line of text meets
Be the spacing mean value Mb that the spacing of above-mentioned adjacent two line of text more approaches the adjacent line of text of large spacing
ave, above-mentioned adjacent two line of text are set to different paragraphs, otherwise above-mentioned adjacent two line of text are set to identical paragraph.
Please refer to Fig. 9 and Figure 10, Fig. 9 is step 105 particular flow sheet in process flow diagram shown in Fig. 1, and Figure 10 is the generation schematic diagram of the word picture after the rearrangement of step 105 of word picture rearrangement method of the present invention.Step 105 comprises:
Step 1052, resets all text column of each paragraph on background picture successively from top to bottom, left to right line by line, the word picture (as shown in figure 10) after being reset.Wherein can be at the width of the first trip indentation user appointment of each paragraph, and certain paragraph line-spacing is set between each paragraph; As the col width of text column to be reset is greater than the residue width width of word blank parts (without) of the line of text of resetting, forward text column to be reset to next line of text and reset processing, guarantee the integrality of text column as far as possible; But the col width of text column is greater than the width of background picture, after text column being taken apart, reset again processing.As excessive in the col width of background row between adjacent two text column, a little gap can be set between these adjacent two text column and replace this background row.In the time of can not having held more line of text as the word picture after current rearrangement, close and preserve the word picture after current rearrangement, a newly-built new background picture is proceeded word rearrangement simultaneously.
The invention still further relates to a kind of word picture system for rearranging, please refer to Figure 11, Figure 11 is the structural representation of the preferred embodiment of word picture system for rearranging of the present invention.This word picture system for rearranging comprises picture modular converter 111, divides module 112, pixel denoising module 113, line of text determination module 114, text column determination module 115, text fragment determination module 116 and reordering module 117.Picture modular converter 111, for being converted to gray scale picture by word picture to be reset; Divide module 112, for all pixels of word picture to be reset being divided into background pixel and text pixel by image binaryzation; Pixel denoising module 113, for carrying out denoising to text pixel; Line of text determination module 114, for according to text pixel, determines the line of text of word picture to be reset; Text column determination module 115, for according to the line of text of text pixel and word picture to be reset, determines the text column of line of text; Text fragment determination module 116, for high according to the spacing of adjacent line of text and row, determines the paragraph at the line of text place of word picture to be reset; Reordering module 117, for according to the paragraph at the size of the word picture after default rearrangement, text column and text column place, the text column for the treatment of in the word picture of rearrangement is reset processing, the word picture after being reset.
This division module 112 comprises poor division unit between gradient division unit and class.Gradient division unit, divides for the pixel that adopts the gradient of pixel grey scale to treat the word picture of rearrangement; Between class, poor division unit, divides for the pixel that adopts the inter-class variance of pixel grey scale to treat the word picture of rearrangement.
Line of text determination module 114 comprises pixel column identifying unit, row denoising unit, preliminary line of text generation unit and line of text generation unit.Pixel column identifying unit, for according to the number of the text pixel of pixel column, judges that text pixel column and background pixel are capable; Row denoising unit, the text pixel column group forming for detection of the text pixel column by continuous up and down, as the number of the text pixel of the text pixel column in text pixel column group is less than Wn
max* Hr
t, text pixel column is set as to background pixel capable; Preliminary line of text generation unit, for adjacent text pixel column is merged to processing, obtains preliminary line of text; Line of text generation unit, for high according to the row of preliminary line of text, merges processing to preliminary line of text, obtains line of text.
Text column determination module 115 comprises pixel column identifying unit, row denoising unit, preliminary text column generation unit and text column generation unit.Pixel column identifying unit, for according to the number of the text pixel of pixel column, judges text pixel column and background pixel row; Row denoising unit, for detection of the text pixel column groups consisting of the right continuous text pixel column in a left side, as the number of the text pixel of the text pixel column in text pixel column groups is less than Wn
max* Wr
t, the background pixel that text pixel column is set as to this article one's own profession is listed as; Preliminary text column generation unit, for adjacent text pixel column is merged to processing, obtains the preliminary text column of line of text; Text column generation unit, for according to the col width of the preliminary text column of line of text, merges processing to preliminary text column, obtains the text column of line of text.
Text fragment determination module 116 comprises paragraph setup unit and the high sudden change of row detecting unit.Paragraph setup unit, for according to the spacing of adjacent line of text, is set the paragraph at line of text place; The high sudden change of row detecting unit, for as high than being greater than the first setting value in the row of adjacent two line of text, or is less than the inverse of the first setting value, adjacent two line of text is set to different paragraphs.
Reordering module 117 comprises background creating unit and word rearrangement unit.Background creating unit, for according to the size of the word picture after default rearrangement, creates a background picture; Unit reset in word, for successively all text column of each paragraph being reset to background picture line by line, and corresponding paragraph marks is set, the word picture after being reset.This word is reset unit and is comprised line feed subelement and take subelement apart.Line feed subelement, for as the col width of text column to be reset be greater than the residue width of the line of text of resetting, forward text column to be reset to next line of text and reset processing; Take subelement apart, for the width of background picture as described in being greater than as the col width of text column, after text column being taken apart, reset again processing.Background creating unit comprises filling subelement, and this fills subelement, for using background pixel to fill background picture.
When word picture system for rearranging of the present invention is used, first judgement needs picture modular converter 111 that word picture to be reset is converted to gray scale picture.Then dividing each pixel that the gradient division unit of module 112 adopts the gradient of pixel grey scale to treat in the word picture of rearrangement divides, as use gradient division unit to divide unsuccessfully, divide poor division unit between the class of module 112 and adopt the inter-class variance of pixel grey scale to divide each pixel in the word picture of resetting.So final division module 112 divides all pixels in word picture to be reset for background pixel and text pixel.The specific works flow process of dividing module 112 refers to the step 101 in the specific embodiment of above-mentioned word picture rearrangement method.
The module of pixel denoising subsequently 113, in order to remove the impact of picture noise, is carried out denoising to the text pixel after dividing.The specific works flow process of pixel denoising module 113 refers to the denoising part of the step 101 in the specific embodiment of above-mentioned word picture rearrangement method.
The pixel column identifying unit of line of text determination module 114, according to the number of the text pixel of pixel column, judges that text pixel column and background pixel are capable subsequently; Row denoising unit is according to the number of the text pixel of pixel column, and the result of determination that pixel column identifying unit is judged is carried out denoising; The adjacent text pixel column of preliminary text column generation unit after to denoising merges processing, obtains preliminary line of text; Line of text generation unit is high according to the row of preliminary line of text, and preliminary line of text is merged to processing, obtains line of text.So, determined the line of text of word picture to be reset.The specific works flow process of line of text determination module 114 refers to the step 102 in the specific embodiment of above-mentioned word picture rearrangement method.
The pixel column identifying unit of text column determination module 115, for according to the number of the text pixel of pixel column, is judged text pixel column and background pixel row subsequently; Row denoising unit is according to the number of the text pixel of pixel column, and the result of determination that pixel column identifying unit is judged is carried out denoising; The adjacent text pixel column of preliminary text column generation unit after to denoising merges processing, obtains the preliminary text column of line of text; Preliminary line of text generation unit, according to the col width of the preliminary text column of line of text, merges processing to preliminary text column, obtains the text column of this article one's own profession.So, determined the text column of the line of text of word picture to be reset.The specific works flow process of text column determination module 115 refers to the step 103 in the specific embodiment of above-mentioned word picture rearrangement method.
The high sudden change of the row detecting unit of text fragment determination module 116 is high by the row of more adjacent line of text line by line subsequently, and the irrelevant word of some and body matter, by the mode of segmentation, is demarcated with body matter word; The word that the paragraph setup unit of text fragment determination module 116 is treated the body matter in the word picture of rearrangement subsequently carries out staging treating.The specific works flow process of text fragment determination module 116 refers to the step 104 in the specific embodiment of above-mentioned word picture rearrangement method.
The background creating unit of reordering module 117, according to the size of the word picture after default rearrangement, creates a background picture subsequently; Then use background pixel to fill background picture; Word is reset unit and successively all text column of each paragraph is reset on background picture from top to bottom, left to right line by line, the word picture after being reset.The specific works flow process of reordering module 117 refers to the step 105 in the specific embodiment of above-mentioned word picture rearrangement method.
Word picture rearrangement method of the present invention and system are reset processing to word picture, and reset word picture and mobile phone screen sizableness after processing, make the word picture of resetting after processing be adapted at browsing on mobile phone, simplify operation when cellphone subscriber reads word, solved the technical matters of mobile phone user word picture inconvenience.
In sum; although the present invention discloses as above with preferred embodiment; but above preferred embodiment is not in order to limit the present invention; those of ordinary skill in the art; without departing from the spirit and scope of the present invention; all can do various changes and retouching, so the scope that protection scope of the present invention defines with claim is as the criterion.
Claims (14)
1. a word picture rearrangement method, is characterized in that, comprising:
By image binaryzation, each pixel in word picture to be reset is divided into background pixel and text pixel;
According to described text pixel, the line of text of word picture to be reset described in determining;
According to described text pixel and described in the line of text of word picture to be reset, determine the text column of described line of text;
High according to the spacing of adjacent described line of text and row, the paragraph at the line of text place of word picture to be reset described in determining; And
According to the paragraph at the size of the word picture after default rearrangement, described text column and described text column place, the text column in word picture described to be reset is reset to processing, obtain the word picture after described rearrangement.
2. word picture rearrangement method according to claim 1, is characterized in that, the described step that all pixels in word picture to be reset is divided into background pixel and text pixel by image binaryzation comprises:
The pixel gradient value of each pixel described in calculating in word picture to be reset;
The mean value of the pixel gradient value of each pixel described in setting in word picture to be reset is pixel gradient mean value, the described pixel that described pixel gradient value is more than or equal to described pixel gradient mean value is set as described text pixel, and the described pixel that described pixel gradient value is less than to described pixel gradient mean value is set as described background pixel.
3. word picture rearrangement method according to claim 1, is characterized in that, the described step that all pixels in word picture to be reset is divided into background pixel and text pixel by image binaryzation comprises:
The grey scale pixel value of choosing variance maximum in each pixel grey scale variance is gray scale segmentation threshold;
According to described gray scale segmentation threshold, all pixels in word picture described to be reset are divided into and are less than or equal to the pixel of described gray scale segmentation threshold and pixel two classes that gray-scale value is greater than described gray scale segmentation threshold, the class pixel that number is many is set as described text pixel, and the class pixel that number is few is set as described background pixel.
4. word picture rearrangement method according to claim 1, is characterized in that, before the described step that by image binaryzation, all pixels in word picture to be reset is divided into background pixel and text pixel, also comprises:
Word picture described to be reset is converted to gray scale picture.
5. word picture rearrangement method according to claim 1, it is characterized in that, described by image binaryzation, all pixels in word picture to be reset are divided into the step of background pixel and text pixel after, described according to described text pixel, described in determining, before the step of the line of text of word picture to be reset, also comprise:
As all neighbors of processed text pixel be as described in background pixel, change described processed text pixel into described background pixel.
6. word picture rearrangement method according to claim 1, is characterized in that, described according to described text pixel, and described in determining, the step of the line of text of word picture to be reset comprises:
By all pixels of word picture to be reset described in line scanning, as be scanned in capable pixel as described in the number of text pixel be greater than 0, will described in be scanned row and be set as text pixel column, otherwise will described in be scanned row to be set as background pixel capable;
Adjacent described text pixel column is merged to processing, obtain preliminary line of text;
High according to the row of described preliminary line of text, described preliminary line of text is merged to processing, obtain described line of text.
7. word picture rearrangement method according to claim 1, is characterized in that, described according to described text pixel and described in the line of text of word picture to be reset, determine that the step of the text column of described line of text comprises:
By the line of text of word picture to be reset described in column scan, as be scanned in the pixel of row as described in the number of text pixel be greater than 0, by described, be scanned the text pixel column that row are set as described line of text, otherwise described being scanned is listed as to the background pixel row that are set as described line of text;
Adjacent described text pixel column is merged to processing, obtain the preliminary text column of described line of text;
According to the col width of the preliminary text column of described line of text, described preliminary text column is merged to processing, obtain the text column of described line of text.
8. word picture rearrangement method according to claim 1, it is characterized in that, described according to the paragraph at size, described text column and the described text column place of the word picture after resetting, text column in word picture described to be reset is reset to processing, and the step that obtains the word picture after described rearrangement comprises:
Size according to the word picture after default rearrangement, creates a background picture;
Successively all text column of each paragraph are reset on described background picture line by line, and corresponding paragraph marks is set, obtain the word picture after described rearrangement.
9. a word picture system for rearranging, is characterized in that, comprising:
Divide module, for each pixel of word picture to be reset being divided into background pixel and text pixel by image binaryzation;
Line of text determination module, for according to described text pixel, the line of text of word picture to be reset described in determining;
Text column determination module, for according to described text pixel and described in the line of text of word picture to be reset, determine the text column of described line of text;
Text fragment determination module, for high according to the spacing of adjacent described line of text and row, the paragraph at the line of text place of word picture to be reset described in determining; And
Reordering module, for according to the paragraph at the size of the word picture after default rearrangement, described text column and described text column place, resets processing to the text column in word picture described to be reset, and obtains the word picture after described rearrangement.
10. word picture system for rearranging according to claim 9, is characterized in that, described division module comprises:
Gradient division unit, for the pixel gradient value of each pixel of word picture to be reset described in calculating;
The mean value of the pixel gradient value of each pixel described in setting in word picture to be reset is pixel gradient mean value, the described pixel that described pixel gradient value is more than or equal to described pixel gradient mean value is set as described text pixel, and the described pixel that described pixel gradient value is less than to described pixel gradient mean value is set as described background pixel; And
Between class, poor division unit, is gray scale segmentation threshold for choosing the grey scale pixel value of each pixel grey scale variance variance maximum;
According to described gray scale segmentation threshold, all pixels in word picture described to be reset are divided into and are less than or equal to the pixel of described gray scale segmentation threshold and pixel two classes that gray-scale value is greater than described gray scale segmentation threshold, the class pixel that number is many is set as described text pixel, and the class pixel that number is few is set as described background pixel.
11. word picture system for rearranging according to claim 9, is characterized in that, described word picture system for rearranging also comprises:
Picture modular converter, for being converted to gray scale picture by word picture described to be reset; And
Pixel denoising module, for described text pixel is carried out to denoising, as all neighbors of processed text pixel be as described in background pixel, change described processed text pixel into described background pixel.
12. word picture system for rearranging according to claim 9, is characterized in that, described line of text determination module comprises:
Pixel column identifying unit, be used for by all pixels of word picture to be reset described in line scanning, as be scanned in capable pixel as described in the number of text pixel be greater than 0, will described in be scanned row and be set as text pixel column, otherwise will described in be scanned row to be set as background pixel capable;
Preliminary line of text generation unit, for adjacent described text pixel column is merged to processing, obtains preliminary line of text; And
Line of text generation unit, high for the row according to described preliminary line of text, described preliminary line of text is merged to processing, obtain described line of text.
13. word picture system for rearranging according to claim 9, is characterized in that, described text column determination module comprises:
Pixel column identifying unit, be used for by the line of text of word picture to be reset described in column scan, as be scanned in the pixel of row as described in the number of text pixel be greater than 0, by described, be scanned the text pixel column that row are set as described line of text, otherwise described being scanned is listed as to the background pixel row that are set as described line of text;
Preliminary text column generation unit, for adjacent described text pixel column is merged to processing, obtains the preliminary text column of described line of text; And
Text column generation unit, for according to the col width of the preliminary text column of described line of text, merges processing to described preliminary text column, obtains the text column of described line of text.
14. word picture system for rearranging according to claim 9, is characterized in that, described reordering module comprises:
Background creating unit, for according to the size of the word picture after default rearrangement, creates a background picture; And
Unit reset in word, for successively all text column of each paragraph being reset to described background picture line by line, and corresponding paragraph marks is set, and obtains the word picture after described rearrangement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210295562.4A CN103593329B (en) | 2012-08-17 | 2012-08-17 | A kind of text image rearrangement method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210295562.4A CN103593329B (en) | 2012-08-17 | 2012-08-17 | A kind of text image rearrangement method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103593329A true CN103593329A (en) | 2014-02-19 |
| CN103593329B CN103593329B (en) | 2018-03-13 |
Family
ID=50083479
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210295562.4A Active CN103593329B (en) | 2012-08-17 | 2012-08-17 | A kind of text image rearrangement method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103593329B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103905837A (en) * | 2014-03-26 | 2014-07-02 | 小米科技有限责任公司 | Image processing method and device and terminal |
| CN104994312A (en) * | 2015-07-15 | 2015-10-21 | 北京金山安全软件有限公司 | Video generation method and device |
| CN106503629A (en) * | 2016-10-10 | 2017-03-15 | 语联网（武汉）信息技术有限公司 | A kind of dictionary picture dividing method and device |
| CN107203764A (en) * | 2016-03-18 | 2017-09-26 | 北大方正集团有限公司 | Long microblog picture recognition methods and device |
| CN108391107A (en) * | 2017-04-01 | 2018-08-10 | 苏州龙诺法智能科技有限公司 | Intelligent translation projecting apparatus controller, control system and control method |
| CN110126496A (en) * | 2019-05-31 | 2019-08-16 | 佛山市科自智能系统技术有限公司 | Concentric loop based on digital ink-jet technology traverses circular image Method of printing |
| CN110941972A (en) * | 2018-09-21 | 2020-03-31 | 广州金山移动科技有限公司 | Method and device for segmenting characters in PDF document and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005063364A (en) * | 2003-08-20 | 2005-03-10 | Toshiba Solutions Corp | Page production system |
| CN101937428A (en) * | 2010-08-11 | 2011-01-05 | 优视科技有限公司 | Method and system for rearranging pictures with literal contents for mobile terminal equipment |
| CN102136064A (en) * | 2011-03-24 | 2011-07-27 | 成都四方信息技术有限公司 | System for recognizing characters from image |
| CN102542279A (en) * | 2010-12-23 | 2012-07-04 | 汉王科技股份有限公司 | Method and device for extracting Uighur, Kazakh and Kirgiz text images by rows |
| CN102567300A (en) * | 2011-12-29 | 2012-07-11 | 方正国际软件有限公司 | Picture document processing method and device |
-
2012
- 2012-08-17 CN CN201210295562.4A patent/CN103593329B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005063364A (en) * | 2003-08-20 | 2005-03-10 | Toshiba Solutions Corp | Page production system |
| CN101937428A (en) * | 2010-08-11 | 2011-01-05 | 优视科技有限公司 | Method and system for rearranging pictures with literal contents for mobile terminal equipment |
| CN102542279A (en) * | 2010-12-23 | 2012-07-04 | 汉王科技股份有限公司 | Method and device for extracting Uighur, Kazakh and Kirgiz text images by rows |
| CN102136064A (en) * | 2011-03-24 | 2011-07-27 | 成都四方信息技术有限公司 | System for recognizing characters from image |
| CN102567300A (en) * | 2011-12-29 | 2012-07-11 | 方正国际软件有限公司 | Picture document processing method and device |
Non-Patent Citations (1)
| Title |
|---|
| 岳宁: "多区域图像的分割和倾斜检测方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103905837A (en) * | 2014-03-26 | 2014-07-02 | 小米科技有限责任公司 | Image processing method and device and terminal |
| CN103905837B (en) * | 2014-03-26 | 2017-04-19 | 小米科技有限责任公司 | Image processing method and device and terminal |
| CN104994312A (en) * | 2015-07-15 | 2015-10-21 | 北京金山安全软件有限公司 | Video generation method and device |
| CN107203764A (en) * | 2016-03-18 | 2017-09-26 | 北大方正集团有限公司 | Long microblog picture recognition methods and device |
| CN107203764B (en) * | 2016-03-18 | 2020-08-07 | 北大方正集团有限公司 | Long microblog picture identification method and device |
| CN106503629A (en) * | 2016-10-10 | 2017-03-15 | 语联网（武汉）信息技术有限公司 | A kind of dictionary picture dividing method and device |
| CN108391107A (en) * | 2017-04-01 | 2018-08-10 | 苏州龙诺法智能科技有限公司 | Intelligent translation projecting apparatus controller, control system and control method |
| CN110941972A (en) * | 2018-09-21 | 2020-03-31 | 广州金山移动科技有限公司 | Method and device for segmenting characters in PDF document and electronic equipment |
| CN110941972B (en) * | 2018-09-21 | 2023-11-28 | 广州金山移动科技有限公司 | Segmentation method and device for characters in PDF document and electronic equipment |
| CN110126496A (en) * | 2019-05-31 | 2019-08-16 | 佛山市科自智能系统技术有限公司 | Concentric loop based on digital ink-jet technology traverses circular image Method of printing |
| CN110126496B (en) * | 2019-05-31 | 2020-10-27 | 佛山市科自智能系统技术有限公司 | Printing method of concentric ring traversal circular image based on digital ink-jet technology |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103593329B (en) | 2018-03-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103593329A (en) | Text image rearrangement method and system | |
| US8218875B2 (en) | Method and system for preprocessing an image for optical character recognition | |
| CN101373465B (en) | Translation apparatus, and translation method | |
| CN106991422B (en) | Character cutting method and device, computer readable storage medium and electronic equipment | |
| CN106313918A (en) | Image processing apparatus and image processing method | |
| US9565338B2 (en) | Image processing apparatus, method, and medium to perform image smoothing and brightness correction to remove show through | |
| CN104103054A (en) | Image processing apparatus and control method thereof | |
| CN103377462A (en) | Method and device for processing scanned image | |
| CN100530219C (en) | Image processing apparatus | |
| US7137558B2 (en) | Methods and apparatus for forming barcode characters | |
| JP2004165969A (en) | Image processing apparatus and program thereof | |
| CN108734846A (en) | Denoising method and device, the terminal and storage medium of banknote image | |
| US20110286672A1 (en) | Translucent image detection apparatus, translucent image edge detection apparatus, translucent image detection method, and translucent image edge detection method | |
| US8472072B2 (en) | Method for improving printing quality and related printing system | |
| US8126193B2 (en) | Image forming apparatus and method of image forming | |
| US20070103704A1 (en) | Methods and apparatuses for merging and outputting images | |
| US9858512B2 (en) | Image processing apparatus with an improved detection of ruled lines | |
| US8368965B2 (en) | Image cropping process | |
| CN112329774B (en) | Commodity ruler code table automatic generation method based on image | |
| JP2012095173A (en) | Image processing device, image formation device, and image processing method | |
| KR100467565B1 (en) | Local binarization method of the imaging system. | |
| JP5453215B2 (en) | Image processing apparatus, image forming apparatus, and image processing method | |
| CN105469362A (en) | Image processing method and image processing device | |
| CN101388951A (en) | Image forming apparatus, image processing apparatus and wire narrowing method | |
| CN102096903B (en) | Page rasterized character smooth processing method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |