CN115578729B - AI intelligent process arrangement method for digital staff - Google Patents
AI intelligent process arrangement method for digital staff Download PDFInfo
- Publication number
- CN115578729B CN115578729B CN202211457579.5A CN202211457579A CN115578729B CN 115578729 B CN115578729 B CN 115578729B CN 202211457579 A CN202211457579 A CN 202211457579A CN 115578729 B CN115578729 B CN 115578729B
- Authority
- CN
- China
- Prior art keywords
- image
- character
- graph
- correction
- original image
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/146—Aligning or centring of the image pick-up or image-field
- G06V30/1475—Inclination or skew detection or correction of characters or of image to be recognised
- G06V30/1478—Inclination or skew detection or correction of characters or of image to be recognised of characters or characters lines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/146—Aligning or centring of the image pick-up or image-field
- G06V30/1463—Orientation detection or correction, e.g. rotation of multiples of 90 degrees
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/16—Image preprocessing
- G06V30/1607—Correcting image deformation, e.g. trapezoidal deformation caused by perspective
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/18—Extraction of features or characteristics of the image
- G06V30/18105—Extraction of features or characteristics of the image related to colour
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Input (AREA)
Abstract
The invention discloses a digital employee AI intelligent process arrangement method, which comprises the following steps: collecting an original image of a paper file with process information, and carrying out difference graying on the original image to obtain a plurality of difference grayed images; rotating the difference gray-scale image by a preset angle to obtain a plurality of rotated gray-scale images; expanding the rotary grayscale image, and detecting a character straight line formed by each line of expanded characters by using Hough transform to obtain a character walking diagram; carrying out perspective transformation on the rotated grayscale image before expansion according to the character walking diagram to obtain a correction diagram; and extracting arrow marks in the correction graph, carrying out affine transformation on the correction graph by taking the arrow marks as auxiliary information, rotating to obtain a restored graph, carrying out binarization on the restored graph, inputting the restored graph to a character recognition module for recognition, and sequentially extracting flow information to complete arrangement. The invention can obtain accurate character trend, avoid recognition error caused by reasons such as special angles and the like, and is beneficial to improving the processing speed and accuracy.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a digital employee AI intelligent process arrangement method.
Background
At present, the flow information recorded in the paper document is inefficient if manually input into a computer, so image recognition is a common solution. Among them, OCR (Optical Character Recognition) refers to a process in which an electronic device checks characters printed on paper, determines a shape thereof by detecting dark and light patterns, and then translates the shape into computer data by a Character Recognition method. The method is a technology for converting characters in a paper document into an image file of a black-and-white dot matrix in an optical mode aiming at print characters, and converting the characters in the image into a text format through recognition software for further editing and processing by character processing software. The main indicators for measuring the performance of an OCR system are: rejection rate, false recognition rate, recognition speed, stability of the product, usability and the like. RPA digital staff performs deep fusion of traditional character recognition and machine learning, can parse data from non-standard documents, and is helpful for converting handwritten text characters into a machine-readable format. In most cases, OCR is mainly used to simplify and convert paper services into digital services, such as: PDF, scanned documents, paper invoices, faxes, handwritten documents, and the like.
However, for the recognition of paper documents, because the document placement may be irregular or the angle may not be aligned, the direction of the actually photographed character may change, and meanwhile, the individual document contains a table or a flowchart, and the conventional character direction judgment cannot be used, the prior art aims at the problem, and particularly, when the angle difference is large, the accurate recognition may be difficult.
Disclosure of Invention
The invention provides a digital staff AI intelligent flow arrangement method aiming at the problem that the angle or direction of characters is difficult to judge in the process of identifying paper documents in the prior art, which mainly aims at the preprocessing of character identification, automatically corrects the angle and direction of the characters, avoids the identification error or failure caused by special tables or flow charts and other reasons, is beneficial to improving the processing speed and accuracy, and obtains accurate and clear results which are convenient for subsequent identification.
The technical scheme of the invention is as follows.
The digital employee AI intelligent process arrangement method comprises the following steps:
s1: collecting an original image of a paper file with process information, and carrying out difference graying on the original image to obtain a plurality of difference grayed images;
s2: rotating the difference gray-scale image by a preset angle to obtain a plurality of rotated gray-scale images;
s3: expanding the rotary grayscale image, and detecting a character straight line formed by each line of expanded characters by using Hough transform to obtain a character walking diagram;
s4: carrying out perspective transformation on the rotated grayscale image before expansion according to the character walking diagram to obtain a correction diagram;
s5: and extracting arrow marks in the correction graph, carrying out affine transformation on the correction graph by taking the arrow marks as auxiliary information, rotating to obtain a restored graph, carrying out binarization on the restored graph, inputting the restored graph to a character recognition module for recognition, and sequentially extracting flow information to complete arrangement.
The invention can prevent the problem of unclear images possibly caused by single graying through differential graying, can ensure that at least one image with a smaller angle with the correct position appears through the rotation of a preset angle so as to reduce the error probability in the subsequent transformation process, and finally can avoid the identification error or the identification failure caused by special tables or flow charts and the like through a series of transformations and by means of the arrow representation identification flow information, thereby being beneficial to improving the speed and the accuracy of processing.
Preferably, the performing the differential graying on the original image includes:
carrying out average graying on the original image by taking the average value of RGB values as a gray value to obtain an average grayed image;
carrying out maximum graying on the original image by taking the maximum value in the RGB values as a gray value to obtain a maximum grayed image;
and carrying out weighted average graying on the original image by combining the RGB value with the preset weight to obtain a weighted average grayed image.
Preferably, the process of obtaining the preset weight includes:
calculating the ratio of pixels with R values larger than a critical value to the total pixels in the original image to obtain a first ratio, calculating the ratio of pixels with G values larger than the critical value to the total pixels to obtain a second ratio, and calculating the ratio of pixels with B values larger than the critical value to the total pixels to obtain a third ratio;
and determining the preset weight of each value of RGB in equal proportion according to the first ratio, the second ratio and the third ratio.
In the scheme, taking the ratio of the pixels with the R values larger than the critical value to the total pixels as an example, the larger the first ratio is, the larger the color ratio of the R values on the whole image is, the larger the influence degree on the image is, so that when the preset weight of each value of RGB is determined in equal proportion, the larger the weight obtained by the R values is, and otherwise, the smaller the ratio is, the smaller the obtained weight is; the method can strengthen the difference caused by the color characteristics of the image, and is particularly suitable for the image processing task of character recognition, because compared with the common image, the color parameters of the character and the background in the image mainly comprising the character are usually obviously broken, and the difference caused by the broken parameter can be amplified by the method. The threshold value is generally set to be about 128 and can be adjusted according to actual needs.
Preferably, the rotating the difference grayed image by a preset angle to obtain a plurality of rotated grayed images includes:
setting preset angles to be-90 degrees, 90 degrees and 180 degrees, sequentially selecting one preset angle for each difference gray-scale image, and rotating to obtain a plurality of rotation gray-scale images. Generally, the image with an uncertain angle and the expected righting angle are most easily recognized when the included angle is smaller than 45 degrees, but in fact, the image may be laid flat or placed upside down, and the recognition difficulty is seriously increased, so that at least one image with the righting angle smaller than 45 degrees can be obtained inevitably through the rotation, the recognition accuracy probability is increased, and the character recognition is facilitated.
Preferably, the obtaining a correction map by subjecting the rotated grayscale image before expansion to perspective transformation according to the character walking diagram includes:
and taking any character straight line in the character walking graph as a reference straight line, and locally stretching or compressing pixels of the rotated gray image before expansion so as to enable the rest character straight lines to be parallel to the reference straight line, thereby obtaining the correction graph.
Preferably, the extracting an arrow mark in the correction map, performing affine transformation on the correction map using the arrow mark as auxiliary information, and rotating to obtain the reduction map includes:
judging the direction of an arrow mark in the same correction chart to obtain a plurality of unit vectors, calculating the total vector of the unit vectors, and judging the pointing direction (x, y) of the total vector;
rotating the correction diagram until the character straight line in the correction diagram is at the horizontal position and y in the pointing direction (x, y) of the total vector is less than or equal to 0 to obtain a candidate diagram;
and screening according to the actual rotation angle of the candidate image relative to the original image, and reserving at least one qualified candidate image as a restoration image.
The scheme performs targeted optimization on the recognition of the flow chart with the arrow, the flow chart generally has a form from top to bottom as a whole, but the directions of the local branch arrows are not consistent, so that the judgment is performed according to the pointing direction of the total vector, when y is less than or equal to 0 after the rotation, the total vector has a downward component, and the condition is met no matter whether the total vector deviates to the left or to the right. This step can filter out the rotated inverted image.
Preferably, the screening according to the actual rotation angle of the candidate map relative to the original image and retaining at least one qualified candidate map as the restoration map includes:
and judging actual rotation angles of different candidate graphs obtained by processing the same original image relative to the original image, calculating the numerical distribution of each actual rotation angle, reserving the actual rotation angles with numerical difference values within 10%, deleting the candidate graphs corresponding to the rest actual rotation angles, and taking the rest candidate graphs as the restored graphs. Although the character line is in a horizontal position, it is not excluded that the image is processed upside down, which may be reduced based on the judgment of the introduced arrow, and this may be substantially removed by further screening.
Preferably, the calculation process of the actual rotation angle includes:
recording a preset angle p of rotation of each rotated grayscale image;
recording a rotation angle q when the correction chart rotates to obtain a candidate chart;
the actual rotation angle C = q + p, wherein clockwise rotation is noted as positive and counter-clockwise rotation is noted as negative.
The substantial effects of the invention include: AI character recognition is carried out on an image to be recognized by digital staff, and a plurality of gray level images with different outstanding color characteristics can be obtained through difference graying, so that the result with the clearest characteristics can be obtained conveniently; at least one image with an included angle smaller than 45 degrees with the correcting angle can be obtained through rotation of a preset angle, and the probability of accurate recognition is increased; the correction process is assisted through integral judgment in the arrow direction; through the progressive steps and the mutual combined action of the steps, the error conditions of angles and directions can be gradually reduced, the correction success rate is increased, the character direction is finally accurately judged, the condition of character inversion cannot occur, and the method is suitable for initial identification of a flow chart.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions will be clearly and completely described below with reference to the embodiments, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be understood that, in various embodiments of the present invention, the sequence numbers of the processes do not mean the execution sequence, and the execution sequence of the processes should be determined by the functions and the internal logic of the processes, and should not constitute any limitation on the implementation process of the embodiments of the present invention.
It should be understood that in the present application, "comprising" and "having" and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be understood that, in the present invention, "a plurality" means two or more. "and/or" is merely an association relationship describing an associated object, meaning that there may be three relationships, for example, and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. "comprises A, B and C" and "comprises A, B, C" means that all three of A, B, C comprise, "comprises A, B or C" means that one of three of A, B, C is comprised, "comprises A, B and/or C" means that any 1 or any 2 or 3 of the three of A, B, C is comprised.
The technical solution of the present invention will be described in detail below with specific examples. Embodiments may be combined with each other and some details of the same or similar concepts or processes may not be repeated in some embodiments.
Example (b):
the digital employee AI intelligent process arrangement method, as shown in fig. 1, includes the following steps:
s1: collecting an original image of a paper file with process information, and carrying out difference graying on the original image to obtain a plurality of difference grayed images. The method comprises the following steps:
carrying out average graying on the original image by taking the average value of RGB values as a gray value to obtain an average grayed image;
carrying out maximum graying on the original image by taking the maximum value in the RGB values as a gray value to obtain a maximum grayed image;
and carrying out weighted average graying on the original image by combining the RGB value with the preset weight to obtain a weighted average grayed image.
Wherein, the process of obtaining the preset weight comprises the following steps:
calculating the ratio of pixels with R values larger than a critical value to the total pixels in the original image to obtain a first ratio, calculating the ratio of pixels with G values larger than the critical value to the total pixels to obtain a second ratio, and calculating the ratio of pixels with B values larger than the critical value to the total pixels to obtain a third ratio;
and determining the preset weight of each value of RGB in equal proportion according to the first ratio, the second ratio and the third ratio.
In the scheme, the ratio of the pixels with the R values larger than the critical value to the total pixels is taken as an example, the larger the first ratio is, the larger the color ratio of the R values on the whole image is, the larger the influence degree on the image is, so that when the preset weight of each value of RGB is determined in equal proportion, the larger the weight obtained by the R values is, and otherwise, the smaller the ratio is, the smaller the obtained weight is; the method can strengthen the difference caused by the color characteristics of the image, and is particularly suitable for the image processing task of character recognition, because compared with the common image, the color parameters of the character and the background in the image mainly comprising the character are usually obviously broken, and the difference caused by the broken parameter can be amplified by the method.
In most cases, the expected effect can be obtained by the conventional grayscaling method, but sometimes it cannot work, for example, in a photographed image, the background is very similar to the text color, the background color is greenish, the RGB value is (180,250,100), which accounts for 70% of the whole image, and the text portion color is bluish, the RGB value is (100,180,250), which accounts for 30% of the whole image. If the average value is grayed, the obtained gray values of the background and the character are the same, which is obviously not beneficial to subsequent identification; if the gray level is maximized, the gray levels of the obtained background and text are still the same, which is obviously not beneficial to the subsequent identification. In most cases, the gradation values obtained by these processing methods are not equal, but in this case, gradation must be performed by another method.
If the weighted average graying of the present embodiment is adopted, in the case where the critical value is 128, the pixels having R values greater than 128 account for 70%, and thus the first ratio is 0.7; the G values of all pixels are greater than 128, so the second ratio is 1; similarly, the third ratio is 0.3. The preset weight of each value of RGB is determined in equal proportion, then the preset weight of R is 0.35, the preset weight of g is 0.5, the preset weight of b is 0.15, and thus, the gray value of the background is 203 and the gray value of the text is 162.5. Because the colors of the original images are very similar, the difference seen by naked eyes after graying is not obvious, but compared with the conventional mode, the difference is obviously distinguished, and a clearer image can be obtained by adjusting the critical value.
S2: and rotating the difference gray-scale image by a preset angle to obtain a plurality of rotated gray-scale images. The method comprises the following steps:
setting preset angles to be-90 degrees, 90 degrees and 180 degrees, sequentially selecting one preset angle for each difference gray-scale image, and rotating to obtain a plurality of rotation gray-scale images. Generally, the image with an uncertain angle and the expected righting angle are most easily recognized when the included angle is smaller than 45 degrees, but in fact, the image may be laid flat or placed upside down, and the recognition difficulty is seriously increased, so that at least one image with the righting angle smaller than 45 degrees can be obtained inevitably through the rotation, the recognition accuracy probability is increased, and the character recognition is facilitated.
S3: and expanding the rotary grayscale image, and detecting a character straight line formed by each line of expanded characters by using Hough transform to obtain a character walking diagram.
The most common method for tilt correction is hough transform, which is based on the principle that an image is expanded to connect discontinuous characters into a straight line, thereby facilitating straight line detection.
S4: and carrying out perspective transformation on the rotated grayscale image before expansion according to the character walking diagram to obtain a corrected diagram. The method comprises the following steps:
and taking any character straight line in the character walking graph as a reference straight line, and locally stretching or compressing pixels of the rotated gray image before expansion so as to enable the rest character straight lines to be parallel to the reference straight line, thereby obtaining the correction graph. The process is similar to the process of trapezoidal correction, and the angle caused by the shooting position can be corrected.
S5: and extracting arrow marks in the correction graph, carrying out affine transformation on the correction graph by taking the arrow marks as auxiliary information, rotating to obtain a restored graph, carrying out binarization on the restored graph, inputting the restored graph to a character recognition module for recognition, and sequentially extracting flow information to complete arrangement. For correcting the oblique picture to the horizontal position. The method comprises the following steps:
judging the direction of an arrow mark in the same correction chart to obtain a plurality of unit vectors, calculating the total vector of the unit vectors, and judging the pointing direction (x, y) of the total vector;
rotating the correction diagram until the character straight line in the correction diagram is at the horizontal position and y in the pointing direction (x, y) of the total vector is less than or equal to 0 to obtain a candidate diagram;
and screening according to the actual rotation angle of the candidate image relative to the original image, and reserving at least one qualified candidate image as a restoration image.
The scheme performs targeted optimization on the recognition of the flow chart with the arrow, the flow chart generally has a form from top to bottom as a whole, but the directions of the locally branched arrows are not consistent, so the judgment is performed according to the pointing direction of the total vector, when y is less than or equal to 0 after rotation, the total vector has a downward component, and the condition is met no matter whether the total vector deviates to the left or to the right. This step can filter out the inverted image after rotation.
In addition, the process of extracting the restored image includes:
and judging actual rotation angles of different candidate graphs obtained by processing the same original image relative to the original image, calculating the numerical distribution of each actual rotation angle, reserving the actual rotation angles with numerical difference values within 10%, deleting the candidate graphs corresponding to the rest actual rotation angles, and taking the rest candidate graphs as the restored graphs.
Wherein the calculation process of the actual rotation angle comprises the following steps:
recording a preset angle p of rotation of each rotated grayscale image;
recording a rotation angle q when the correction chart rotates to obtain a candidate chart;
the actual rotation angle C = q + p, wherein clockwise rotation is noted as positive and counter-clockwise rotation is noted as negative.
Although the character line is in a horizontal position, it is not excluded that the image is processed upside down, which may be reduced based on the judgment of the introduced arrow, and this may be substantially removed by further screening.
It should be noted that through the above steps, the error conditions of the angle and the direction can be gradually reduced, the correction success rate is increased, the character direction is finally accurately judged, the situation of character inversion does not occur, and the method is suitable for the initial recognition of the flow chart. The steps realize the result that 1+1 is greater than 2, and the lack of any step can cause other steps to lose the optimal effect, thus leading to inaccurate result.
The embodiment can prevent the problem of unclear images possibly caused by single graying through differential graying, can ensure that at least one image with a smaller angle with the correct position appears through the rotation of a preset angle so as to reduce the error probability in the subsequent transformation process, finally represents identification flow information through a series of transformations and by means of an arrow, does not generate the situation of character inversion, can avoid identification errors or identification failures caused by special tables, flow charts and other reasons, and is favorable for improving the processing speed and accuracy.
In the embodiments provided in this application, it should be understood that the disclosed structures and methods may be implemented in other ways, or some features may be omitted, or not implemented.
In addition, the embodiments of the present application may be integrated into one processing unit, or each of the units may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially or partially contributed to by the prior art, or all or part of the technical solutions may be embodied in the form of a software product, where the software product is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a variety of media that can store program codes, such as a usb disk, a removable hard disk, a Read Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (4)
1. The digital employee AI intelligent process arrangement method is characterized by comprising the following steps:
s1: collecting an original image of a paper file with process information, and carrying out difference graying on the original image to obtain a plurality of difference grayed images;
s2: rotating the difference gray-scale image by a preset angle to obtain a plurality of rotated gray-scale images;
s3: expanding the rotary grayscale image, and detecting a character straight line formed by each line of expanded characters by using Hough transform to obtain a character walking diagram;
s4: carrying out perspective transformation on the rotated grayscale image before expansion according to the character walking diagram to obtain a correction diagram;
s5: extracting arrow marks in the correction graph, performing affine transformation on the correction graph by taking the arrow marks as auxiliary information, rotating to obtain a restored graph, inputting the restored graph after binaryzation into a character recognition module for recognition, and sequentially extracting flow information to complete arrangement;
the differential graying of the original image comprises the following steps:
carrying out average graying on the original image by taking the average value of RGB values as a gray value to obtain an average grayed image;
carrying out maximum graying on the original image by taking the maximum value in the RGB values as a gray value to obtain a maximum grayed image;
carrying out weighted average graying on the original image by combining RGB values with preset weight to obtain a weighted average grayed image;
the method for extracting the arrow mark in the correction graph, performing affine transformation on the correction graph by taking the arrow mark as auxiliary information, and rotating to obtain the restored graph comprises the following steps:
judging the direction of an arrow mark in the same correction chart to obtain a plurality of unit vectors, calculating the total vector of the unit vectors, and judging the pointing direction (x, y) of the total vector;
rotating the correction diagram until the character straight line in the correction diagram is at the horizontal position and y in the pointing direction (x, y) of the total vector is less than or equal to 0 to obtain a candidate diagram;
screening according to the actual rotation angle of the candidate image relative to the original image, and reserving at least one qualified candidate image as a restoration image; the screening according to the actual rotation angle of the candidate image relative to the original image and reserving at least one qualified candidate image as a restoration image comprises the following steps:
judging actual rotation angles of different candidate graphs obtained by processing the same original image relative to the original image, calculating the numerical distribution of each actual rotation angle, reserving the actual rotation angles with numerical difference values within 10%, deleting the candidate graphs corresponding to the rest actual rotation angles, and taking the rest candidate graphs as restored graphs;
the calculation process of the actual rotation angle includes:
recording a preset angle p of rotation of each rotated grayscale image;
recording a rotation angle q when the correction chart rotates to obtain a candidate chart;
the actual rotation angle C = q + p, wherein clockwise rotation is noted as positive and counter-clockwise rotation is noted as negative.
2. The digital employee AI intelligent process orchestration method of claim 1, wherein the process of obtaining the preset weights comprises:
calculating the ratio of pixels with R values larger than a critical value to the total pixels in the original image to obtain a first ratio, calculating the ratio of pixels with G values larger than the critical value to the total pixels to obtain a second ratio, and calculating the ratio of pixels with B values larger than the critical value to the total pixels to obtain a third ratio;
and determining the preset weight of each value of RGB in equal proportion according to the first ratio, the second ratio and the third ratio.
3. The AI intelligent process arrangement method for digital employees according to claim 1, wherein said rotating the difference grayed images by a preset angle to obtain a plurality of rotated grayed images comprises:
setting preset angles to be-90 degrees, 90 degrees and 180 degrees, sequentially selecting one preset angle for each difference gray-scale image, and rotating to obtain a plurality of rotation gray-scale images.
4. The AI intelligent process arrangement method for digital staff according to claim 1, wherein said perspective transformation of the rotated grayed images before expansion according to the character walking diagram to obtain a correctional diagram comprises:
and taking any character straight line in the character walking graph as a reference straight line, and locally stretching or compressing pixels of the rotated gray image before expansion so as to enable the rest character straight lines to be parallel to the reference straight line, thereby obtaining the correction graph.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211457579.5A CN115578729B (en) | 2022-11-21 | 2022-11-21 | AI intelligent process arrangement method for digital staff |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211457579.5A CN115578729B (en) | 2022-11-21 | 2022-11-21 | AI intelligent process arrangement method for digital staff |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115578729A CN115578729A (en) | 2023-01-06 |
| CN115578729B true CN115578729B (en) | 2023-03-21 |
Family
ID=84588185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211457579.5A Active CN115578729B (en) | 2022-11-21 | 2022-11-21 | AI intelligent process arrangement method for digital staff |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115578729B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113160222A (en) * | 2021-05-14 | 2021-07-23 | 电子科技大学 | Production data identification method for industrial information image |
| WO2021179485A1 (en) * | 2020-03-11 | 2021-09-16 | 平安科技（深圳）有限公司 | Image rectification processing method and apparatus, storage medium, and computer device |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6317223B1 (en) * | 1998-12-14 | 2001-11-13 | Eastman Kodak Company | Image processing system for reducing vertically disposed patterns on images produced by scanning |
| CN114787934A (en) * | 2019-11-25 | 2022-07-22 | 通用电气精准医疗有限责任公司 | Algorithm orchestration of workflows to facilitate healthcare imaging diagnostics |
| CN114926839B (en) * | 2022-07-22 | 2022-10-14 | 富璟科技（深圳）有限公司 | Image identification method based on RPA and AI and electronic equipment |
-
2022
- 2022-11-21 CN CN202211457579.5A patent/CN115578729B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021179485A1 (en) * | 2020-03-11 | 2021-09-16 | 平安科技（深圳）有限公司 | Image rectification processing method and apparatus, storage medium, and computer device |
| CN113160222A (en) * | 2021-05-14 | 2021-07-23 | 电子科技大学 | Production data identification method for industrial information image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115578729A (en) | 2023-01-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109657665B (en) | Invoice batch automatic identification system based on deep learning | |
| US6778703B1 (en) | Form recognition using reference areas | |
| US7978900B2 (en) | Systems for mobile image capture and processing of checks | |
| CN109086714A (en) | Table recognition method, identifying system and computer installation | |
| CN111476109A (en) | Bill processing method, bill processing apparatus, and computer-readable storage medium | |
| CN112183038A (en) | Form identification and typing method, computer equipment and computer readable storage medium | |
| CN110598566A (en) | Image processing method, device, terminal and computer readable storage medium | |
| CN113903024A (en) | Handwritten bill numerical value information identification method, system, medium and device | |
| JPWO2017141802A1 (en) | Image processing apparatus, character recognition apparatus, image processing method, and program recording medium | |
| CN111626145A (en) | Simple and effective incomplete form identification and page-crossing splicing method | |
| CN114266764A (en) | Character integrity detection method and device for printed label | |
| CN108197624A (en) | The recognition methods of certificate image rectification and device, computer storage media | |
| CN112419207A (en) | Image correction method, device and system | |
| CN110210467B (en) | Formula positioning method of text image, image processing device and storage medium | |
| JP4013060B2 (en) | Image correction method and image correction apparatus | |
| JP3099771B2 (en) | Character recognition method and apparatus, and recording medium storing character recognition program | |
| CN111814780B (en) | Bill image processing method, device, equipment and storage medium | |
| CN115578729B (en) | AI intelligent process arrangement method for digital staff | |
| US6694059B1 (en) | Robustness enhancement and evaluation of image information extraction | |
| CN112597868A (en) | Test paper identification and correction method based on no positioning point | |
| EP0651337A1 (en) | Object recognizing method, its apparatus, and image processing method and its apparatus | |
| CN116110069A (en) | Answer sheet identification method and device based on coding mark points and relevant medium thereof | |
| CN113537216B (en) | Dot matrix font text line inclination correction method and device | |
| CN113837119A (en) | Method and equipment for recognizing confusable characters based on gray level images | |
| US6351560B1 (en) | Apparatus, method, and program for locating an objective on a form and a recording medium for recording the program for locating the objective on the form |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |