CN114119410B - Method and device for correcting cells in distorted tabular image - Google Patents
Method and device for correcting cells in distorted tabular image Download PDFInfo
- Publication number
- CN114119410B CN114119410B CN202111401956.9A CN202111401956A CN114119410B CN 114119410 B CN114119410 B CN 114119410B CN 202111401956 A CN202111401956 A CN 202111401956A CN 114119410 B CN114119410 B CN 114119410B
- Authority
- CN
- China
- Prior art keywords
- offset
- distortion
- distorted
- cell
- frame
- 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
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012937 correction Methods 0.000 claims abstract description 95
- 238000001514 detection method Methods 0.000 claims abstract description 85
- 238000003709 image segmentation Methods 0.000 claims abstract description 35
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims abstract description 10
- 230000009466 transformation Effects 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000013500 data storage Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 238000004590 computer program Methods 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000007175 bidirectional communication Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G06T5/80—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
Abstract
The present disclosure relates to a method and apparatus for correcting cells in a distorted table image, the method for correcting cells in a distorted table image comprising: acquiring a distortion table image; carrying out image segmentation on the distortion table image to obtain each cell in the distortion table image; determining distorted cells in the respective cells; carrying out target detection on the distorted table image to obtain a distorted unit cell detection frame; calculating the intersection and parallel ratio of the distortion unit cell and the corresponding distortion unit cell detection frame to determine the distortion degree of the distortion unit cell; determining a frame line for correcting the distortion cell and a correction frame for correcting contents within the frame line based on the distortion degree; affine transformation is performed on the frame lines of the distortion unit cells and the contents in the frame lines based on the correction frames. According to the present disclosure, the frame lines of the cells in the distorted form image and the contents within the frame lines can be corrected.
Description
Technical Field
The present disclosure relates to the field of image processing technologies, and in particular, to a method and an apparatus for correcting cells in a distorted table image.
Background
Currently, the most common sources of form images are: the document is obtained by scanning the document with a scanner or by shooting the document with a camera of the electronic apparatus or the like. However, in actual practice, it is inevitable that the scanned or photographed form image is distorted. When it is desired to perform character recognition (OCR) on content in a spreadsheet image, the accuracy of recognition is greatly affected if there is distortion in the image. Therefore, it is necessary to correct the form image before performing character recognition on the form image to facilitate the character recognition.
Disclosure of Invention
The present disclosure provides a method and apparatus for correcting cells in a distorted form image, an electronic device, and a computer-readable storage medium to solve at least the problems of the related art described above.
According to a first aspect of embodiments of the present disclosure, there is provided a method of correcting a cell in a distorted table image, comprising: acquiring a distortion table image; carrying out image segmentation on the distortion table image to obtain each cell in the distortion table image; determining distorted cells in the respective cells; carrying out target detection on the distorted table image to obtain a distorted unit cell detection frame; calculating the intersection and parallel ratio of the distortion unit cell and the corresponding distortion unit cell detection frame to determine the distortion degree of the distortion unit cell; determining a frame line for correcting the distortion cell and a correction frame for correcting contents within the frame line based on the distortion degree; affine transformation is performed on the frame lines of the distortion unit cells and the contents in the frame lines based on the correction frames.
Optionally, the determining a correction frame for correcting the outline of the distorted cell and the content in the outline based on the distortion degree includes: carrying out approximate fitting on the frame lines of the distorted cells to obtain the minimum circumscribed rectangle of the distorted cells; and determining a middle area correction frame of the minimum bounding rectangle and the distorted cell detection frame as the correction frame based on the distortion degree.
Optionally, the diagonal coordinates [ x1, y1, x2, y2] of the middle region correction box are calculated by: [ x1, y1, x2, y2] ([ xa1, ya1, xa2, ya2] + [ xb1, yb1, xb2, yb2 ]). times.D/2, where [ xa1, ya1, xa2, ya2] are diagonal coordinates of the minimum bounding rectangle, [ xb1, yb1, xb2, yb2] are diagonal coordinates of the distorted cell detection box, D is the distortion degree, and D is 0.5. ltoreq. D.ltoreq.1.
Optionally, the determining a correction frame for correcting the outline of the distorted cell and the content in the outline based on the distortion degree includes: determining a distorted position of the distorted cell; and carrying out concave-convex correction on the distortion position based on the distortion degree to obtain the correction frame.
Alternatively, the coordinates of the pixel point of the distortion position after the irregularity correction are [ xd ± offset _ x, yd ± offset _ y ], where [ xd, yd ] is the coordinates of the pixel point of the distortion position before the irregularity correction, offset _ x is the correction amount for xd, offset _ y is the correction amount for yd, offset _ x is 0, offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)) in the case where the distortion position is located only on the lateral frame line of the distortion cell, and when yd is corrected in the manner of yd + offset _ y, if yd + offset _ y > yp, offset _ y is 0, when yd is corrected in the manner of yd-offset _ y, if yd-offset _ y < y, offset _ y is 0, offset _ y is located only on the longitudinal frame line of the distortion cell, offset _ y is 0, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), and when xd + offset _ x is used to correct xd, if xd + offset _ x > xp, offset _ x is 0, when xd is used to correct xd, if xd-offset _ x < xp, offset _ x is 0, and when the distortion position is located on both the lateral and vertical frame lines, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when xd + offset _ x is used to correct xd, if xd + offset _ x > xd, offset _ x is 0, when xd + offset _ x is used to correct xd, if xd-offset _ x < xp, offset _ x is 0, if yd + offset _ y > yp, if yd + offset _ y is corrected, offset _ y is 0, and if yd-offset _ y < yp, offset _ y is 0, where H is the height of the distorted cell detection frame, W is the width of the distorted cell detection frame, D is the distortion degree, 0.5 or less D is less than 1, xp is the abscissa of the undistorted position on the same vertical frame line as [ xd, yd ], and yp is the ordinate of the undistorted position on the same horizontal frame line as [ xd, yd ], when yd is corrected by yd-offset _ y, if yd-offset _ y is smaller than yp.
Optionally, the determining a distorted cell of the respective cells includes: respectively extracting the features of each cell; determining the distortion cell from the extracted features.
Optionally, the image segmentation is performed on the distorted table image through a trained image segmentation model, so as to obtain each cell in the distorted table image.
Optionally, target detection is performed on the distorted table image through a trained target detection model, so as to obtain a distorted cell detection frame.
Optionally, before the image segmentation and the target detection are performed on the distorted table image, the method further includes: converting the distortion table image into a gray image to obtain a gray image of the distortion table image; and carrying out binarization processing on the distorted table image gray-scale map to obtain a binarized distorted table image.
According to a second aspect of the embodiments of the present disclosure, there is provided an apparatus for correcting a cell in a distorted form image, including: an image acquisition unit configured to acquire a distortion table image; an image segmentation unit configured to perform image segmentation on the distortion table image to obtain each cell in the distortion table image; a determination unit configured to determine a distorted cell among the respective cells; the target detection unit is configured to perform target detection on the distorted table image to obtain a distorted cell detection frame; a calculation unit configured to calculate an intersection ratio of the distortion cell to the corresponding distortion cell detection frame to determine a distortion degree of the distortion cell; a correction frame determination unit configured to determine a frame line for correcting the distortion cell and a correction frame for correcting contents within the frame line based on the degree of distortion; an affine transformation unit configured to perform affine transformation on a frame line of the distortion cell and contents within the frame line based on the correction frame.
Optionally, the correction frame determination unit is configured to: carrying out approximate fitting on the frame lines of the distorted cells to obtain the minimum circumscribed rectangle of the distorted cells; and determining a middle area correction frame of the minimum bounding rectangle and the distorted cell detection frame as the correction frame based on the distortion degree.
Optionally, the diagonal coordinates [ x1, y1, x2, y2] of the middle region correction box are calculated by: [ x1, y1, x2, y2] ([ xa1, ya1, xa2, ya2] + [ xb1, yb1, xb2, yb2 ]). times.D/2, where [ xa1, ya1, xa2, ya2] are diagonal coordinates of the minimum bounding rectangle, [ xb1, yb1, xb2, yb2] are diagonal coordinates of the distorted cell detection box, D is the distortion degree, and D is 0.5. ltoreq. D.ltoreq.1.
Optionally, the correction frame determination unit is configured to: determining a distorted position of the distorted cell; and carrying out concave-convex correction on the distortion position based on the distortion degree to obtain the correction frame.
Alternatively, the coordinates of the pixel point of the distortion position after the irregularity correction are [ xd ± offset _ x, yd ± offset _ y ], where [ xd, yd ] is the coordinates of the pixel point of the distortion position before the irregularity correction, offset _ x is the correction amount for xd, offset _ y is the correction amount for yd, offset _ x is 0, offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)) in the case where the distortion position is located only on the lateral frame line of the distortion cell, and when yd is corrected in the manner of yd + offset _ y, if yd + offset _ y > yp, offset _ y is 0, when yd is corrected in the manner of yd-offset _ y, if yd-offset _ y < y, offset _ y is 0, offset _ y is located only on the longitudinal frame line of the distortion cell, offset _ y is 0, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), and when xd + offset _ x is used to correct xd, if xd + offset _ x > xp, offset _ x is 0, when xd is used to correct xd, if xd-offset _ x < xp, offset _ x is 0, and when the distortion position is located on both the lateral and vertical frame lines, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when xd + offset _ x is used to correct xd, if xd + offset _ x > xd, offset _ x is 0, when xd + offset _ x is used to correct xd, if xd-offset _ x < xp, offset _ x is 0, if yd + offset _ y > yp, if yd + offset _ y is corrected, offset _ y is 0, and if yd-offset _ y < yp, offset _ y is 0, where H is the height of the distorted cell detection frame, W is the width of the distorted cell detection frame, D is the distortion degree, 0.5 or less D is less than 1, xp is the abscissa of the undistorted position on the same vertical frame line as [ xd, yd ], and yp is the ordinate of the undistorted position on the same horizontal frame line as [ xd, yd ], when yd is corrected by yd-offset _ y, if yd-offset _ y is smaller than yp.
Optionally, the determining unit is configured to: respectively extracting the features of each cell; determining the distortion cell from the extracted features.
Optionally, the image segmentation unit is configured to: and carrying out image segmentation on the distorted table image through a trained image segmentation model to obtain each cell in the distorted table image.
Optionally, the target detection unit is configured to: and carrying out target detection on the distorted table image through a trained target detection model to obtain a distorted unit cell detection frame.
Optionally, the image acquisition unit is further configured to: after the distortion table image is obtained, converting the distortion table image into a gray image to obtain a gray image of the distortion table image; and carrying out binarization processing on the distorted table image gray-scale map to obtain a binarized distorted table image.
According to a third aspect of the embodiments of the present disclosure, there is provided an electronic apparatus including: a processor; a memory for storing computer executable instructions; wherein the computer executable instructions, when executed by the processor, cause the processor to implement the method as described above.
According to a fourth aspect of embodiments of the present disclosure, there is provided a computer-readable storage medium, wherein instructions, when executed by a processor, cause the processor to perform the method as described above.
The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:
according to the method and the device for correcting the cells in the distorted table image, the frame lines of the cells in the distorted table image and the content in the frame lines can be corrected so as to be beneficial to character recognition.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure and are not to be construed as limiting the disclosure.
Fig. 1 is an implementation scenario diagram illustrating a method of correcting cells in a distorted form image according to an exemplary embodiment of the present disclosure.
Fig. 2 is a flowchart illustrating a method of correcting cells in a distorted form image according to an exemplary embodiment of the present disclosure.
Fig. 3 is a schematic diagram for explaining the cross-over ratio.
Fig. 4 is a block diagram illustrating an apparatus for correcting a cell in a distorted table image according to an exemplary embodiment of the present disclosure.
Fig. 5 is a block diagram illustrating an electronic device according to an exemplary embodiment of the present disclosure.
Detailed Description
In order to make the technical solutions of the present disclosure better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.
It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The embodiments described in the following examples do not represent all embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.
In this case, the expression "at least one of the items" in the present disclosure means a case where three types of parallel expressions "any one of the items", "a combination of any plural ones of the items", and "the entirety of the items" are included. For example, "include at least one of a and B" includes the following three cases in parallel: (1) comprises A; (2) comprises B; (3) including a and B. For another example, "at least one of the first step and the second step is performed", which means that the following three cases are juxtaposed: (1) executing the step one; (2) executing the step two; (3) and executing the step one and the step two.
At present, some distortion inevitably exists in a form image acquired by scanning or photographing. This greatly affects the accuracy of subsequent text recognition of the form image.
Therefore, the present disclosure provides a method and an apparatus for correcting cells in a distorted table image, which can correct the frame lines of the cells in the distorted table image and the content (characters, etc.) in the frame lines, thereby improving the accuracy of subsequent character recognition.
Fig. 1 is an implementation scenario diagram illustrating a method of correcting cells in a distorted form image according to an exemplary embodiment of the present disclosure. As shown in fig. 1, the implementation scenario includes a server 100, user terminals 110 and 120. The user terminals 110 and 120 and the server 100 may be used to implement the method for correcting cells in the distorted form image of the present disclosure, and the correction may be performed by the server 100 after the distorted form image is acquired by the user terminals 110 and 120. At this time, the server 100 can perform bidirectional communication with the user terminals 110 and 120. The user terminals are not limited to the number and types shown in the figures, and include mobile terminals such as smart phones and tablet computers with camera functions, personal computers, and the like, and may also include any other electronic devices capable of shooting or scanning. The server 100 may be a single server, a server cluster composed of several servers, a cloud computing platform, or a virtualization center.
Next, a method and apparatus for correcting cells in a distorted form image according to an exemplary embodiment of the present disclosure will be described with reference to fig. 2 to 4.
Fig. 2 is a flowchart illustrating a method of correcting cells in a distorted form image according to an exemplary embodiment of the present disclosure.
As shown in fig. 2, in step S210, a distortion table image is acquired. Here, the distorted table image is an image in which distortion occurs in a normal table image without any distortion. The distortion table image may be acquired by various methods in the related art. Also, the distorted form image may be an image that is distorted due to an improper operation when a document with a form is scanned or photographed. The distortion may be a case where a part of the form image is deformed without being in a normal position.
Then, in step S220, the distortion table image acquired in step S210 is subjected to image segmentation to obtain each cell in the distortion table image. Here, each cell in the distortion table image can be obtained by various conventional image segmentation methods. According to an exemplary embodiment of the disclosure, the distorted table image may be subjected to image segmentation by a trained image segmentation model, resulting in each cell in the distorted table image. Here, the trained image segmentation model is obtained by training in advance based on the target of image segmentation. This makes it possible to accurately obtain each cell in the distortion table image. Further, the distortion table image may be preprocessed as described below prior to image segmentation of the distortion table image to facilitate image segmentation.
According to an exemplary embodiment of the present disclosure, before step S220, the method may further include: converting the distorted table image into a gray image to obtain a gray image of the distorted table image; and then, carrying out binarization processing on the distorted table image gray-scale map to obtain a binarized distorted table image. That is, in the case where the distorted table image is not a gray map, it is converted into a gray map, and then the gray values of the pixel points in the gray map are set to 0 or 255 (for thresholding, 255 is set when less than the threshold value and 0 is set when greater than the threshold value), thereby obtaining a binarized (and inverted) distorted table image. Here, the distortion table image may be converted into a grayscale image by various existing methods. The binarization processing may be performed by various conventional methods. For example, a binarized image that can reflect the entire image and local features can be obtained by appropriate threshold selection. By performing such preprocessing on the distorted table image, subsequent image segmentation and target detection can be facilitated.
Then, in step S230, a distortion cell among the respective cells obtained by image division is determined. Here, the distorted cell is a cell with distortion. The cells with distortion in each cell can be determined by various existing methods.
According to an exemplary embodiment of the present disclosure, step S230 may include: feature extraction is performed on each cell, and then a distorted cell is determined according to the extracted features. For example, whether a cell is a distorted cell may be determined based on a feature that can indicate that the cell is distorted, such as a frame line feature of the cell. The feature may be extracted by various existing methods, or may be extracted by a predetermined feature extraction model.
Then, in step S240, target detection is performed on the distortion table image acquired in step S210, resulting in a distortion cell detection frame. Here, the target of the target detection is a distorted cell. Similarly, the distorted table image may be preprocessed before the target detection is performed on the distorted table image, so as to facilitate the target detection. The specific method of pretreatment is as described above. Here, the distorted cell detection frame can be detected by various existing object detection methods. According to the exemplary embodiment of the disclosure, the distorted table image can be subjected to target detection through a trained target detection model, so that a distorted cell detection frame is obtained. Here, the trained object detection model is obtained by training in advance according to an object to be detected. The target detection model can adopt a Yolov3 network, and the Yolov3 network comprises: darkent53 (feature extractor model), FPN (feature pyramid network), Yolo Head function, etc. In the present invention, the size of the distortion cell detection frame is taken as the size of the cell of the distortion cell in the normal case for determining the distortion degree. The order of step S240 and step S220 may be changed or may be executed simultaneously, and this does not affect the execution of the present invention.
Then, in step S250, the intersection ratio of the distortion cell obtained by image division and the corresponding distortion cell detection frame is calculated to determine the distortion degree of the distortion cell. Here, the Intersection over ratio (IoU) is the ratio of the Intersection area of the distortion cell and the distortion cell detection box to the phase area (as shown in fig. 3), and is used to measure the similarity between the two. The intersection ratio has a value range between 0 and 1, wherein 0 means that the two do not have overlapped pixels, and 1 means that the two are equal. In the present invention, the intersection ratio and the distortion degree are linearly corresponding, and the larger the intersection ratio is, the smaller the distortion degree is, and the smaller the intersection ratio is, the larger the distortion degree is. Here, the distortion degree of the distortion cell may be determined based on the cross-over ratio by various existing methods.
Then, in step S260, based on the distortion degree, a frame line for correcting the distortion cell and a correction frame for correcting the content within the frame line are determined.
According to an exemplary embodiment of the present disclosure, step S260 may include: and performing approximate fitting on the frame lines of the distorted cells to obtain the minimum circumscribed rectangle of the distorted cells. Then, based on the distortion degree, a middle area correction frame of the minimum bounding rectangle and the distortion cell detection frame is determined as a correction frame. That is, the middle area between the minimum bounding rectangle and the distorted cell detection frame is taken as the correction frame. Here, the minimum bounding rectangle may be obtained by an existing approximation fitting method.
According to an exemplary embodiment of the present disclosure, diagonal coordinates [ x1, y1, x2, y2] of the middle region correction box of the minimum bounding rectangle and the distorted cell detection box may be calculated by:
[x1，y1，x2，y2]＝([xa1，ya1，xa2，ya2]+[xb1，yb1，xb2，yb2])×D/2，
wherein [ xa1, ya1, xa2 and ya2] are diagonal coordinates of the minimum circumscribed rectangle, [ xb1, yb1, xb2 and yb2] are diagonal coordinates of the distorted cell detection box, D is the distortion degree, and D is more than or equal to 0.5 and less than or equal to 1. This makes it possible to obtain a frame line for correcting the distortion cell and a correction frame for correcting the content in the frame line.
Further, according to an exemplary embodiment of the present disclosure, step S260 may include: the distorted position of the distorted cell is determined. Then, based on the degree of distortion, the distortion position is corrected for unevenness, and a correction frame is obtained. Here, the distortion position of the distortion cell may be determined by various existing methods. Then, the correction frame can be obtained by performing convex correction on the distorted position such as a concave and concave correction on the distorted position such as a convex based on the degree of distortion.
According to an exemplary embodiment of the present disclosure, coordinates of a pixel point of a distortion position after concave-convex correction may be set as [ xd ± offset _ x, yd ± offset _ y ], where [ xd, yd ] is coordinates of a pixel point of a distortion position before concave-convex correction, offset _ x is a correction amount for xd, and offset _ y is a correction amount for yd. At this time, when the distortion position is located only on the horizontal frame line of the distortion cell, only the ordinate is corrected, offset _ x is 0, and offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when yd is corrected in the manner of yd + offset _ y (that is, when the distortion position is downward distorted), if yd + offset _ y > yp, offset _ y is 0, and when yd is corrected in the manner of yd-offset _ y (that is, when the distortion position is upward distorted), if yd-offset _ y < yp, offset _ y is 0.
In addition, when the distortion position is located only on the vertical frame line of the distortion cell, only the abscissa is corrected, offset _ y is 0, and offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), and when xd is corrected in the manner of xd + offset _ x (that is, when the distortion position is distorted leftward), if xd + offset _ x > xp, offset _ x is 0, and when xd is corrected in the manner of xd-offset _ x (that is, when the distortion position is distorted rightward), if xd-offset _ x < xp, offset _ x is 0.
Further, in the case where the distortion position is located on both the horizontal frame line and the vertical frame line (i.e., the intersection of the horizontal frame line and the vertical frame line), the abscissa and the ordinate are corrected, respectively, with offset _ x being 1/D × 100 × sin (2 × pi × xd/(2 × W)), offset _ y being 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when xd + offset _ x is used to correct xd (i.e., the distortion position is distorted to the left), if xd + offset _ x > xp, offset _ x is 0, when xd is corrected in the manner of xd-offset _ x (i.e., the distortion position is distorted to the right), if xd-offset _ x < xp, offset _ x is 0, when yd is corrected in the manner of yd + offset _ x, i.e., the distortion position is distorted to the right, when yd-offset _ x < xp, i.e., the distortion position is corrected to the lower direction, if yd + offset _ y > yp, then offset _ y is 0, and when yd is corrected in the manner of yd-offset _ y (i.e., when the distortion position is upward distortion), if yd-offset _ y < yp, then offset _ y is 0. H is the height of the distorted cell detection frame, W is the width of the distorted cell detection frame, D is the distortion degree, and D is more than or equal to 0.5 and less than or equal to 1. Here, xp is the abscissa of the undistorted position on the same vertical frame line as [ xd, yd ]. In addition, a distorted position and an undistorted position exist on the vertical frame line where distortion occurs, and when the distorted position is corrected, excessive correction can be restricted according to the abscissa of the undistorted position on the vertical frame line. yp is the ordinate of the undistorted position on the same transverse frame line as [ xd, yd ]. Similarly, a distorted position and an undistorted position exist on the distorted horizontal frame line, and when the distorted position is corrected, the excessive correction can be restricted according to the ordinate of the undistorted position on the horizontal frame line.
This makes it possible to obtain a frame line for correcting the distortion cell and a correction frame for correcting the content in the frame line.
Then, in step S270, affine transformation is performed on the frame lines of the distortion cells and the contents inside the frame lines based on the correction frame obtained as described above. Thereby, the frame line of the distorted cell and the content inside the frame line are corrected.
According to the method for correcting the cells in the distorted table image, the frame lines and the contents in the frame lines of the cells in the distorted table image can be corrected to be beneficial to character recognition.
Fig. 4 is a block diagram illustrating an apparatus for correcting a cell in a distorted table image according to an exemplary embodiment of the present disclosure.
As shown in fig. 4, the apparatus 400 for correcting cells in a distorted form image includes: an image acquisition unit 410, an image segmentation unit 420, a determination unit 430, an object detection unit 440, a calculation unit 450, a correction frame determination unit 460, and an affine transformation unit 470.
Wherein the image acquisition unit 410 is configured to acquire a distortion table image.
According to an exemplary embodiment of the present disclosure, the image acquisition unit 410 may be further configured to: after the distorted table image is obtained, preprocessing is carried out on the distorted table image, the distorted table image is converted into a gray-scale image, and the gray-scale image of the distorted table image is obtained. And then carrying out binarization processing on the distorted table image gray-scale map to obtain a binarized distorted table image. Thereby, subsequent image segmentation and object detection is facilitated.
The image segmentation unit 420 is configured to perform image segmentation on the distortion table image obtained by the image acquisition unit 410, resulting in individual cells in the distortion table image.
According to an exemplary embodiment of the present disclosure, the image segmentation unit 420 may be configured to: and carrying out image segmentation on the distorted table image through the trained image segmentation model to obtain each cell in the distorted table image.
The determination unit 430 is configured to determine a distorted cell among the respective cells into which the image division unit 420 performs the image division.
According to an exemplary embodiment of the present disclosure, the determining unit 430 may be configured to perform feature extraction on the respective cells, and then determine a distorted cell according to the extracted features.
The target detection unit 440 is configured to perform target detection on the distorted table image obtained by the image obtaining unit 410, resulting in a distorted cell detection box.
According to an exemplary embodiment of the present disclosure, the target detection unit 440 may be configured to: and carrying out target detection on the distorted table image through the trained target detection model to obtain a distorted unit cell detection frame.
The calculation unit 450 is configured to calculate an intersection ratio of the distortion cell and the corresponding distortion cell detection frame to determine a distortion degree of the distortion cell.
The correction frame determination unit 460 is configured to determine a correction frame for correcting the outline of the distortion cell and the content within the outline, based on the degree of distortion.
According to an exemplary embodiment of the present disclosure, the correction frame determination unit 460 may be configured to: and performing approximate fitting on the frame lines of the distorted cells to obtain the minimum circumscribed rectangle of the distorted cells. Then, based on the distortion degree, a middle area correction frame of the minimum bounding rectangle and the distortion cell detection frame is determined as a correction frame.
According to an exemplary embodiment of the present disclosure, the diagonal coordinates [ x1, y1, x2, y2] of the middle region correction box are calculated by: [ x1, y1, x2, y2] ([ xa1, ya1, xa2, ya2] + [ xb1, yb1, xb2, yb2 ]). times.D/2, where [ xa1, ya1, xa2, ya2] are the diagonal coordinates of the minimum bounding rectangle, [ xb1, yb1, xb2, yb2] are the diagonal coordinates of the distorted cell detection box, D is the distortion degree, and D is 0.5. ltoreq. D.ltoreq.1. This makes it possible to obtain a frame line for correcting the distortion cell and a correction frame for correcting the content in the frame line.
Further, according to an exemplary embodiment of the present disclosure, the correction frame determination unit 460 may be configured to: the distorted position of the distorted cell is determined. Then, based on the degree of distortion, the distortion position is corrected for unevenness, and a correction frame is obtained.
According to the exemplary embodiment of the present disclosure, coordinates of a pixel point of a distortion position after concave-convex correction are [ xd ± offset _ x, yd ± offset _ y ], where [ xd, yd ] is coordinates of a pixel point of a distortion position before concave-convex correction, offset _ x is a correction amount for xd, offset _ y is a correction amount for yd, and if the distortion position is located only on a lateral frame line of a distortion cell, offset _ x is 0, offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and, when yd is corrected in a manner of yd + offset _ y, if yd + offset _ y > yp, offset _ y is 0, and if yd-offset _ y is corrected in a manner of yd-offset _ y, offset _ y is 0, if yd-offset _ y < yp, offset _ y is located only on a longitudinal frame line, offset _ y is 0, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), and when xd is corrected in the manner of xd + offset _ x, if xd + offset _ x > xp, offset _ x is 0, when xd is corrected in the manner of xd-offset _ x, if xd-offset _ x < xp, offset _ x is 0, when the distortion position is both on the lateral frame line and on the vertical frame line, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when xd is corrected in the manner of xd + offset _ x, if xd + offset _ x > x 100 × sin (2 × pi × yd/(2 × H)), if x + offset _ x > 0, if xd + offset _ x > x _ x, offset _ x is 0, when xd is corrected in the manner of xd + offset _ x, when yd is corrected by using yd + offset _ y, if yd + offset _ y > yp, offset _ y is 0, when yd is corrected by using yd-offset _ y, if yd-offset _ y < yp, offset _ y is 0, wherein H is the height of the distorted cell detection frame, W is the width of the distorted cell detection frame, D is the distortion degree, 0.5 is equal to or less than 1, xp is the abscissa of the undistorted position on the same vertical frame line as [ xd, yd ], and yp is the ordinate of the undistorted position on the same horizontal frame line as [ xd, yd ].
This makes it possible to obtain a frame line for correcting the distortion cell and a correction frame for correcting the content in the frame line.
The affine transformation unit 470 is configured to affine-transform the frame line of the distortion cell and the contents within the frame line based on the above-described correction frame. Thereby, the frame line of the distorted cell and the content inside the frame line are corrected.
According to the device for correcting the cells in the distorted table image, the frame lines and the contents in the frame lines of the cells in the distorted table image can be corrected to be beneficial to character recognition.
It should be understood that the apparatus for correcting cells in a distorted form image according to an exemplary embodiment of the present disclosure may perform the method described above with reference to fig. 2, and thus, in order to avoid redundancy, detailed description thereof is omitted.
Fig. 5 shows a block diagram of an electronic device according to an exemplary embodiment of the present disclosure.
The electronic device 500 includes at least one memory 510 having stored therein a set of computer-executable instructions that, when executed by the at least one processor, cause the at least one processor 520 to perform a method of correcting cells in a distorted form image according to an exemplary embodiment of the disclosure.
By way of example, the electronic device 500 may be a PC computer, tablet device, personal digital assistant, smartphone, or other device capable of executing the set of instructions described above. Here, the electronic device 500 need not be a single electronic device, but can be any collection of devices or circuits that can execute the above instructions (or sets of instructions) individually or in combination. The electronic device 500 may also be part of an integrated control system or system manager, or may be configured as a portable electronic device that interfaces with local or remote (e.g., via wireless transmission).
In the electronic device 500, the processor 520 may include a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a programmable logic device, a special purpose processor system, a microcontroller, or a microprocessor. By way of example, and not limitation, processor 520 may also include an analog processor, a digital processor, a microprocessor, a multi-core processor, a processor array, a network processor, or the like.
In addition, the electronic device 500 may also include a video display (such as a liquid crystal display) and a user interaction interface (such as a keyboard, mouse, touch input device, etc.). All components of the electronic device may be connected to each other via a bus and/or a network.
According to an exemplary embodiment of the present disclosure, there may also be provided a computer-readable storage medium, wherein instructions in the computer-readable storage medium, when executed by at least one processor, cause the at least one processor to perform the method of correcting a cell in a distorted form image of the present disclosure. Examples of the computer-readable storage medium herein include: read-only memory (ROM), random-access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random-access memory (RAM), dynamic random-access memory (DRAM), static random-access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD + R, CD-RW, CD + RW, DVD-ROM, DVD-R, DVD + R, DVD-RW, DVD + RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or compact disc memory, Hard Disk Drive (HDD), solid-state drive (SSD), card-type memory (such as a multimedia card, a Secure Digital (SD) card or a extreme digital (XD) card), magnetic tape, a floppy disk, a magneto-optical data storage device, an optical data storage device, a hard disk, a magnetic tape, a magneto-optical data storage device, a hard disk, a magnetic tape, a magnetic data storage device, a magnetic tape, a magnetic data storage device, a magnetic tape, a magnetic data storage device, a magnetic tape, a magnetic data storage device, a magnetic tape, a magnetic data storage device, A solid state disk, and any other device configured to store and provide a computer program and any associated data, data files, and data structures to a processor or computer in a non-transitory manner such that the processor or computer can execute the computer program. The computer program in the computer-readable storage medium described above can be run in an environment deployed in a computer apparatus, such as a client, a host, a proxy device, a server, and the like, and further, in one example, the computer program and any associated data, data files, and data structures are distributed across a networked computer system such that the computer program and any associated data, data files, and data structures are stored, accessed, and executed in a distributed fashion by one or more processors or computers.
According to an exemplary embodiment of the present disclosure, a computer program product is provided, comprising computer instructions which, when executed by a processor, implement the method of correcting cells in a distorted form image of an exemplary embodiment of the present disclosure.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims (20)
1. A method of correcting cells in a distorted tabular image, comprising:
acquiring a distortion table image;
carrying out image segmentation on the distortion table image to obtain each cell in the distortion table image;
determining distorted cells in the respective cells;
carrying out target detection on the distorted table image to obtain a distorted unit cell detection frame;
calculating the intersection and parallel ratio of the distortion unit cell and the corresponding distortion unit cell detection frame to determine the distortion degree of the distortion unit cell;
determining a frame line for correcting the distortion cell and a correction frame for correcting contents within the frame line based on the distortion degree;
affine transformation is performed on the frame lines of the distortion unit cells and the contents in the frame lines based on the correction frames.
2. The method of claim 1, wherein determining a correction box for correcting a box line of the distorted cell and contents within the box line based on the distortion degree comprises:
carrying out approximate fitting on the frame lines of the distorted cells to obtain the minimum circumscribed rectangle of the distorted cells;
and determining a middle area correction frame of the minimum bounding rectangle and the distorted cell detection frame as the correction frame based on the distortion degree.
3. The method of claim 2, wherein the diagonal coordinates [ x1, y1, x2, y2] of the middle region correction box are calculated by:
[x1，y1，x2，y2]＝([xa1，ya1，xa2，ya2]+[xb1，yb1，xb2，yb2])×D/2，
wherein [ xa1, ya1, xa2 and ya2] are diagonal coordinates of the minimum circumscribed rectangle, [ xb1, yb1, xb2 and yb2] are diagonal coordinates of the distorted cell detection box, D is the distortion degree, and D is more than or equal to 0.5 and less than or equal to 1.
4. The method of claim 1, wherein determining a correction box for correcting a box line of the distorted cell and contents within the box line based on the distortion degree comprises:
determining a distorted position of the distorted cell;
and carrying out concave-convex correction on the distortion position based on the distortion degree to obtain the correction frame.
5. The method according to claim 4, wherein coordinates of a pixel point of the distorted position after the irregularity correction are [ xd ± offset _ x, yd ± offset _ y ], where [ xd, yd ] is coordinates of a pixel point of the distorted position before the irregularity correction, offset _ x is a correction amount for xd, and offset _ y is a correction amount for yd,
in the case where the distortion position is located only on the lateral frame line of the distortion cell, offset _ x is 0, and offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when yd is corrected in the manner of yd + offset _ y, if yd + offset _ y > yp, offset _ y is 0, and when yd is corrected in the manner of yd-offset _ y, if yd-offset _ y < yp, offset _ y is 0,
in the case where the distortion position is located only on the vertical frame line of the distortion cell, offset _ y is 0, and offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), and when xd + offset _ x is used for correction of xd, if xd + offset _ x > xp, offset _ x is 0, and when xd is used for correction of xd in the xd-offset _ x manner, if xd-offset _ x < xp, offset _ x is 0,
in the case where the distortion position is located on both the lateral frame line and the longitudinal frame line, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when xd is corrected in the manner of xd + offset _ x, if xd + offset _ x > xp, offset _ x is 0, and when xd is corrected in the manner of xd-offset _ x, if xd-offset _ x < xp, offset _ x is 0,
when yd is corrected in the manner of yd + offset _ y, if yd + offset _ y > yp, then offset _ y is 0, when yd is corrected in the manner of yd-offset _ y, if yd-offset _ y < yp, then offset _ y is 0,
h is the height of the distorted cell detection frame, W is the width of the distorted cell detection frame, D is the distortion degree, D is more than or equal to 0.5 and less than or equal to 1, xp is the abscissa of the undistorted position on the same longitudinal frame line with [ xd, yd ], yp is the ordinate of the undistorted position on the same transverse frame line with [ xd, yd ].
6. The method of claim 1, wherein determining the distorted cell of the individual cells comprises:
respectively extracting the features of each cell;
determining the distortion cell from the extracted features.
7. The method of claim 1,
and carrying out image segmentation on the distorted table image through a trained image segmentation model to obtain each cell in the distorted table image.
8. The method of claim 1,
and carrying out target detection on the distorted table image through a trained target detection model to obtain a distorted unit cell detection frame.
9. The method of claim 1, wherein prior to image segmentation and target detection of the distortion table image, further comprising:
converting the distortion table image into a gray image to obtain a gray image of the distortion table image;
and carrying out binarization processing on the distorted table image gray-scale map to obtain a binarized distorted table image.
10. An apparatus for correcting cells in a distorted tabular image, comprising:
an image acquisition unit configured to acquire a distortion table image;
an image segmentation unit configured to perform image segmentation on the distortion table image to obtain each cell in the distortion table image;
a determination unit configured to determine a distorted cell among the respective cells;
the target detection unit is configured to perform target detection on the distorted table image to obtain a distorted cell detection frame;
a calculation unit configured to calculate an intersection ratio of the distortion cell to the corresponding distortion cell detection frame to determine a distortion degree of the distortion cell;
a correction frame determination unit configured to determine a frame line for correcting the distortion cell and a correction frame for correcting contents within the frame line based on the degree of distortion;
an affine transformation unit configured to perform affine transformation on a frame line of the distortion cell and contents within the frame line based on the correction frame.
11. The apparatus according to claim 10, wherein the correction frame determination unit is configured to:
carrying out approximate fitting on the frame lines of the distorted cells to obtain the minimum circumscribed rectangle of the distorted cells;
and determining a middle area correction frame of the minimum bounding rectangle and the distorted cell detection frame as the correction frame based on the distortion degree.
12. The apparatus of claim 11, wherein the diagonal coordinates [ x1, y1, x2, y2] of the middle region correction box are calculated by:
[x1，y1，x2，y2]＝([xa1，ya1，xa2，ya2]+[xb1，yb1，xb2，yb2])×D/2，
wherein [ xa1, ya1, xa2 and ya2] are diagonal coordinates of the minimum circumscribed rectangle, [ xb1, yb1, xb2 and yb2] are diagonal coordinates of the distorted cell detection box, D is the distortion degree, and D is more than or equal to 0.5 and less than or equal to 1.
13. The apparatus according to claim 10, wherein the correction frame determination unit is configured to:
determining a distorted position of the distorted cell;
and carrying out concave-convex correction on the distortion position based on the distortion degree to obtain the correction frame.
14. The apparatus according to claim 13, wherein coordinates of a pixel point of the distorted position after the irregularity correction are [ xd ± offset _ x, yd ± offset _ y ], where [ xd, yd ] is coordinates of a pixel point of the distorted position before the irregularity correction, offset _ x is a correction amount for xd, and offset _ y is a correction amount for yd,
in the case where the distortion position is located only on the lateral frame line of the distortion cell, offset _ x is 0, and offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when yd is corrected in the manner of yd + offset _ y, if yd + offset _ y > yp, offset _ y is 0, and when yd is corrected in the manner of yd-offset _ y, if yd-offset _ y < yp, offset _ y is 0,
in the case where the distortion position is located only on the vertical frame line of the distortion cell, offset _ y is 0, and offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), and when xd + offset _ x is used for correction of xd, if xd + offset _ x > xp, offset _ x is 0, and when xd is used for correction of xd in the xd-offset _ x manner, if xd-offset _ x < xp, offset _ x is 0,
in the case where the distortion position is located on both the lateral frame line and the longitudinal frame line, offset _ x is 1/D × 100 × sin (2 × pi × xd/(2 × W)), offset _ y is 1/D × 100 × sin (2 × pi × yd/(2 × H)), and when xd is corrected in the manner of xd + offset _ x, if xd + offset _ x > xp, offset _ x is 0, and when xd is corrected in the manner of xd-offset _ x, if xd-offset _ x < xp, offset _ x is 0,
when yd is corrected in the manner of yd + offset _ y, if yd + offset _ y > yp, then offset _ y is 0, when yd is corrected in the manner of yd-offset _ y, if yd-offset _ y < yp, then offset _ y is 0,
h is the height of the distorted cell detection frame, W is the width of the distorted cell detection frame, D is the distortion degree, D is more than or equal to 0.5 and less than or equal to 1, xp is the abscissa of the undistorted position on the same longitudinal frame line with [ xd, yd ], yp is the ordinate of the undistorted position on the same transverse frame line with [ xd, yd ].
15. The apparatus of claim 10, wherein the determining unit is configured to:
respectively extracting the features of each cell;
determining the distortion cell from the extracted features.
16. The apparatus of claim 10, wherein the image segmentation unit is configured to:
and carrying out image segmentation on the distorted table image through a trained image segmentation model to obtain each cell in the distorted table image.
17. The apparatus of claim 10, wherein the target detection unit is configured to:
and carrying out target detection on the distorted table image through a trained target detection model to obtain a distorted unit cell detection frame.
18. The apparatus of claim 10, wherein the image acquisition unit is further configured to: after the distortion table image is acquired,
converting the distortion table image into a gray image to obtain a gray image of the distortion table image;
and carrying out binarization processing on the distorted table image gray-scale map to obtain a binarized distorted table image.
19. An electronic device, comprising:
a processor;
a memory for storing computer executable instructions;
wherein the computer executable instructions, when executed by the processor, cause the processor to implement the method of any one of claims 1 to 9.
20. A computer-readable storage medium, wherein instructions in the computer-readable storage medium, when executed by a processor, cause the processor to perform the method of any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111401956.9A CN114119410B (en) | 2021-11-19 | 2021-11-19 | Method and device for correcting cells in distorted tabular image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111401956.9A CN114119410B (en) | 2021-11-19 | 2021-11-19 | Method and device for correcting cells in distorted tabular image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114119410A CN114119410A (en) | 2022-03-01 |
| CN114119410B true CN114119410B (en) | 2022-04-22 |
Family
ID=80371612
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111401956.9A Active CN114119410B (en) | 2021-11-19 | 2021-11-19 | Method and device for correcting cells in distorted tabular image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114119410B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115082939B (en) * | 2022-05-12 | 2023-11-17 | 吉林省吉林祥云信息技术有限公司 | System and method for correcting distortion table in image based on arc differentiation |
| CN115937010B (en) * | 2022-08-17 | 2023-10-27 | 北京字跳网络技术有限公司 | Image processing method, device, equipment and medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108510445A (en) * | 2018-03-30 | 2018-09-07 | 长沙全度影像科技有限公司 | A kind of Panorama Mosaic method |
| CN110163198A (en) * | 2018-09-27 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of Table recognition method for reconstructing, device and storage medium |
| CN111931864A (en) * | 2020-09-17 | 2020-11-13 | 南京甄视智能科技有限公司 | Method and system for multiple optimization of target detector based on vertex distance and cross-over ratio |
| CN111989692A (en) * | 2019-09-30 | 2020-11-24 | 北京市商汤科技开发有限公司 | Form recognition method, form extraction method and related device |
| CN112733855A (en) * | 2020-12-30 | 2021-04-30 | 科大讯飞股份有限公司 | Table structuring method, table recovery equipment and device with storage function |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107249096B (en) * | 2016-06-14 | 2021-02-26 | 杭州海康威视数字技术股份有限公司 | Panoramic camera and shooting method thereof |
-
2021
- 2021-11-19 CN CN202111401956.9A patent/CN114119410B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108510445A (en) * | 2018-03-30 | 2018-09-07 | 长沙全度影像科技有限公司 | A kind of Panorama Mosaic method |
| CN110163198A (en) * | 2018-09-27 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of Table recognition method for reconstructing, device and storage medium |
| CN111989692A (en) * | 2019-09-30 | 2020-11-24 | 北京市商汤科技开发有限公司 | Form recognition method, form extraction method and related device |
| CN111931864A (en) * | 2020-09-17 | 2020-11-13 | 南京甄视智能科技有限公司 | Method and system for multiple optimization of target detector based on vertex distance and cross-over ratio |
| CN112733855A (en) * | 2020-12-30 | 2021-04-30 | 科大讯飞股份有限公司 | Table structuring method, table recovery equipment and device with storage function |
Non-Patent Citations (5)
| Title |
|---|
| Fully automatic stitching and distortion correction of transmission electron microscope images;Verena Kaynig 等;《Journal of Structural Biology》;20100505;163-173 * |
| RDCFace: Radial Distortion Correction for Face Recognition;He Zhao 等;《2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)》;20200805;7718-7727 * |
| 一种新的基于规则点阵的图像畸变校正方法;李欢 等;《仪器仪表学报》;20040831;第25卷(第4期);581-582 * |
| 基于卷积网络的受电弓图像目标检测与矫正方法研究;庞鸿宇 等;《电气化铁道》;20211015;1-6 * |
| 基于改进的Faster R-CNN的通用目标检测框架;马佳良 等;《计算机应用》;20210910;第21卷(第9期);2712-2719 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114119410A (en) | 2022-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114119410B (en) | Method and device for correcting cells in distorted tabular image | |
| US10867171B1 (en) | Systems and methods for machine learning based content extraction from document images | |
| CN108416902B (en) | Real-time object identification method and device based on difference identification | |
| US11816710B2 (en) | Identifying key-value pairs in documents | |
| AU2017232186A1 (en) | Fast and robust image alignment for burst mode | |
| US9082181B2 (en) | Image processing method and apparatus | |
| CN110874618B (en) | OCR template learning method and device based on small sample, electronic equipment and medium | |
| US11710210B1 (en) | Machine-learning for enhanced machine reading of non-ideal capture conditions | |
| WO2020060632A1 (en) | Converting an image into a structured table | |
| WO2020125062A1 (en) | Image fusion method and related device | |
| RU2581786C1 (en) | Determination of image transformations to increase quality of optical character recognition | |
| CN107305682B (en) | Method and device for splicing images | |
| CN113436222A (en) | Image processing method, image processing apparatus, electronic device, and storage medium | |
| WO2021168703A1 (en) | Character processing and identifying methods, storage medium, and terminal device | |
| CN112419207A (en) | Image correction method, device and system | |
| US10354161B2 (en) | Detecting font size in a digital image | |
| CN112052702A (en) | Method and device for identifying two-dimensional code | |
| CN114359932B (en) | Text detection method, text recognition method and device | |
| CN115223166A (en) | Picture pre-labeling method, picture labeling method and device, and electronic equipment | |
| CN112102203A (en) | Image correction method, device and equipment | |
| CN108304840B (en) | Image data processing method and device | |
| CN115797291A (en) | Circuit terminal identification method and device, computer equipment and storage medium | |
| CN113780294B (en) | Text character segmentation method and device | |
| CN111476090A (en) | Watermark identification method and device | |
| CN114648751A (en) | Method, device, terminal and storage medium for processing video subtitles |
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 |