CN115049837A - Characteristic diagram interference removing method and screen shot watermark identification method comprising same - Google Patents
Characteristic diagram interference removing method and screen shot watermark identification method comprising same Download PDFInfo
- Publication number
- CN115049837A CN115049837A CN202210963058.0A CN202210963058A CN115049837A CN 115049837 A CN115049837 A CN 115049837A CN 202210963058 A CN202210963058 A CN 202210963058A CN 115049837 A CN115049837 A CN 115049837A
- Authority
- CN
- China
- Prior art keywords
- image
- watermark
- feature
- foreground
- steps
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/0021—Image watermarking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- 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
- G06V30/148—Segmentation of character regions
- G06V30/155—Removing patterns interfering with the pattern to be recognised, such as ruled lines or underlines
-
- 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
- G06V30/162—Quantising the image signal
-
- 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
-
- 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
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/19173—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2201/00—General purpose image data processing
- G06T2201/005—Image watermarking
- G06T2201/0065—Extraction of an embedded watermark; Reliable detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Drying Of Semiconductors (AREA)
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
The invention particularly relates to a characteristic diagram interference removing method and a screen shot watermark identification method comprising the same, wherein the characteristic diagram interference removing method comprises the following steps: carrying out self-adaptive binarization processing on an original image of the feature map to obtain a foreground map and taking the foreground map as a mask image; according to the mask image, etching operation is carried out from outside to inside step by step with a fixed step length; comparing the result after the etching operation with the mask image of the previous stage to obtain an interpolation region; calculating the mean value filtering of the characteristic diagram interpolation area and filling the calculation result into the interpolation area; and continuing the steps until the mask image is corroded to disappear, and obtaining the characteristic diagram after the interference is removed after multiple times of filling. The method can conveniently realize the desalination of the foreground image and effectively remove the interference of the foreground image from the characteristic image.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a feature map interference removing method and a screen shot watermark identification method comprising the same.
Background
Image processing technology is a very popular technology at present, and information that we want can be acquired from an image through a series of processes on the image. In various processing methods, the features in the image can be abstracted by performing operations such as convolution and the like on the image to obtain a feature map, and more complex tasks such as target identification, classification and the like can be automatically realized through the abstracted feature map.
For example, by adding an invisible watermark to a screen, when someone shoots or records screen content and transmits the content to the screen to cause secret leakage, the invisible watermark can be extracted from the leaked shot screen image or recorded video, and the invisible watermark contains user information of the screen, so that the secret leakage can be traced accurately. The invisible watermarks have very good visual effects and are basically invisible to naked eyes, so that when the invisible watermarks are extracted, the images to be processed need to be processed to obtain a feature map, and corresponding information is extracted from the feature map. The invisible watermarks generally cover the whole screen and are fused with the background of the screen, and when the characteristic extraction is performed, the invisible watermarks are found to often interfere with a characteristic diagram displayed on the screen, wherein the influence of characters is most obvious, so that the subsequent watermark extraction is not accurate enough, and therefore, a method capable of removing the interference of the characteristic diagram is urgently needed, and the influence of the content displayed on the screen on the characteristic diagram is reduced.
Due to the particularity of the screen shot scene, the screen area generally needs to be identified, then the screen area is cut out, and is corrected, perspective transformed and the like, and finally the invisible watermark can be extracted, the invisible watermark extraction consumes little time, but the pretreatment process of the screen shot image consumes much time; when the shot screen is not complete enough, errors can be generated in the identification and correction of the screen area, and therefore the invisible watermark extraction fails. Therefore, there is still a need for a method for identifying a screen shot watermark, which can quickly identify whether an invisible watermark exists in a screen shot image.
Disclosure of Invention
The invention aims to provide a method for removing interference of a characteristic diagram, which can conveniently remove the interference of a foreground diagram in the characteristic diagram.
In order to realize the purpose, the invention adopts the technical scheme that: a method for removing characteristic diagram interference comprises the following steps: carrying out self-adaptive binarization processing on an original image of the feature map to obtain a foreground map and taking the foreground map as a mask image; according to the mask image, etching operation is carried out from outside to inside step by step with a fixed step length; comparing the result after the etching operation with the mask image of the previous stage to obtain an interpolation region; calculating the mean value filtering of the characteristic diagram interpolation area and filling the calculation result into the interpolation area; and continuing the steps until the mask image is corroded to disappear, and obtaining the characteristic diagram after the interference is removed after multiple times of filling.
Compared with the prior art, the invention has the following technical effects: the mask image is obtained according to the foreground image, the area needing to be processed can be conveniently obtained, the mask image is corroded from outside to inside according to a fixed step length, the corroded area is filled with mean value filtering, and the process is repeated until the corrosion of the mask image disappears, so that the foreground image can be conveniently faded, and the interference of the foreground image is effectively removed from the characteristic image.
Another object of the present invention is to provide a method for identifying a screen shot watermark, which can accurately identify watermark information of a feature map after removing interference.
In order to realize the purpose, the invention adopts the technical scheme that: a screen shot watermark identification method comprises the following steps: clipping and/or zooming the shot screen image to obtain an original image; extracting the watermark of the original image by using a watermark extraction program to obtain a characteristic image; processing the original image and the characteristic diagram according to the steps to obtain the characteristic diagram after interference removal; constructing a characteristic filter corresponding to the characteristic size of the watermark; extracting and strengthening the feature map after the interference is removed by using a feature filter to obtain a response image; and calculating a watermark possibility score S according to the response image, comparing the S with a set threshold value T, judging that the screen shot image to be detected contains the watermark if the S is larger than the T, and otherwise, judging that the screen shot image to be detected does not contain the watermark.
Compared with the prior art, the invention has the following technical effects: the method comprises the steps of extracting watermarks of a screen shot image to be detected through a watermark extraction program to obtain watermark features, extracting and strengthening the features after interference is removed through a constructed feature filter, calculating probability scores of the features, and comparing the probability scores with a set threshold value to judge whether the features contain the watermarks.
Drawings
FIG. 1 is a flow chart of a signature interference removal method;
FIG. 2 is a screen shot image to be detected;
FIG. 3 is a text mask corresponding to FIG. 2;
FIG. 4 is a comparison of signatures before and after removing interference;
FIG. 5 is a flow chart of a screen shot watermark identification method;
FIG. 6 is a feature filter visualization image;
fig. 7 is a response image of a screen shot image with an invisible watermark.
Detailed Description
The present invention will be described in further detail with reference to fig. 1 to 7.
Referring to fig. 1 to 4, the present invention discloses a method for removing interference of a characteristic diagram, including the following steps: carrying out self-adaptive binarization processing on an original image of the feature map to obtain a foreground map and taking the foreground map as a mask image; according to the mask image, etching operation is carried out from outside to inside step by step with a fixed step length; comparing the result after the etching operation with the mask image of the previous stage to obtain an interpolation region; calculating the mean value filtering of the characteristic diagram interpolation area and filling the calculation result into the interpolation area; and continuing the steps until the mask image is corroded to disappear, and obtaining the characteristic diagram after the interference is removed after multiple times of filling. The mask image is obtained according to the foreground image, the area needing to be processed can be conveniently obtained, the mask image is corroded from outside to inside according to a fixed step length, the corroded area is filled with mean value filtering, and the process is repeated until the corrosion of the mask image disappears, so that the foreground image can be conveniently faded, and the interference of the foreground image is effectively removed from the characteristic image.
Further, to ensure that the foreground map can be sufficiently eliminated, the mask image is set slightly larger than the foreground map. And in the step of carrying out self-adaptive binarization processing on the characteristic image to obtain a foreground image and taking the foreground image as a mask image, taking an image obtained by carrying out expansion processing on the foreground image as the mask image. After the setting, the area included by the mask image is slightly larger than the area of the foreground image, the foreground image can be completely eliminated through subsequent processing, and the mask image is expanded by 1-3 pixels generally.
Further, the fixed step size in the etching operation performed from the outside to the inside step by step is 1-4 pixels. Therefore, the foreground image can be eliminated step by step in a finer manner, if the length and width of the whole image are larger, the step size can be set to be larger, otherwise, more time is consumed, for example, when the image size is 500 × 400 pixels and the step size is 1 pixel when the etching operation is performed in practice, and if the image is larger, for example, 2000 × 1600 pixels, the step size can be set to be 4 pixels correspondingly.
Further, the step of performing adaptive binarization processing on the original image of the feature map to obtain a foreground map, and taking the foreground map as a mask image, wherein the feature map and the original image are obtained according to the following steps: cutting and/or scaling the shot screen image to obtain an original image, wherein the size of the original image is the same as the input size of the watermark extraction program; and (4) carrying out watermark extraction on the original image by using a watermark extraction program to obtain a characteristic image. Since the watermark extraction program generally has a requirement on the size of the original image, the screen image shot first is cropped, scaled, or cropped plus scaled to make the size of the screen image meet the requirement of the watermark extraction program. The watermark extraction program only comprises a watermark extraction function and does not comprise the preprocessing of the image and the subsequent processing of the feature map.
In a preferred embodiment of the present invention, the captured screen image includes text, and as shown in fig. 1, the text area is a foreground image. As can be seen from the upper diagram of fig. 4, the influence of the text on the watermark feature is very large, which is very unfavorable for extracting the watermark feature, and after the processing, the influence caused by the text is basically eliminated from the lower diagram of fig. 4.
Referring to fig. 5, the invention also discloses a screen shot watermark identification method, which comprises the following steps: clipping and/or zooming the shot screen image to obtain an original image; extracting the watermark of the original image by using a watermark extraction program to obtain a characteristic image; processing the original image and the characteristic diagram according to the steps to obtain the characteristic diagram after interference removal; constructing a characteristic filter corresponding to the characteristic size of the watermark; extracting and strengthening the feature map after the interference is removed by using a feature filter to obtain a response image; and calculating a watermark possibility score S according to the response image, comparing the S with a set threshold value T, judging that the screen shot image to be detected contains the watermark if the S is larger than the T, and otherwise, judging that the screen shot image to be detected does not contain the watermark. The method comprises the steps of extracting watermarks of a screen shot image to be detected through a watermark extraction program to obtain watermark characteristics, extracting and strengthening the characteristics after interference removal through a constructed characteristic filter, calculating probability scores of the characteristics, and comparing the probability scores with a set threshold value to judge whether the watermark exists. The cropping and/or scaling, as used herein, includes three cases: cropping, scaling, cropping and scaling, which are very simple image processing steps that are much faster than the steps of perspective transformation, screen area recognition, etc.
Referring to fig. 6 and 7, further, the step of constructing a feature filter corresponding to the feature size of the watermark includes the following steps: the possible structure of the block-shaped watermark is determined according to the block-shaped watermark, the possible structure of the block-shaped watermark is determined according to the shape characteristic of the block-shaped watermark, and the feature filter designed according to the structure can well enhance the watermark characteristic in the feature map.
Taking block watermarks as an example, possible structures include corner structures, straight-edge structures, and checkerboard structures; for corner structures, the shape may be: one corner is a watermark area, and the other three are non-watermark areas; it is also possible that: one corner is a non-watermark region, the other three are watermark regions, and the corresponding are 8 small images in the first row and the second row in fig. 6. For a straight-sided structure, the shape is: half of which are watermark regions and the other half of which are non-watermark regions, corresponding to the 4 mini-maps in the third row of fig. 6. For a checkerboard structure, the shape is only possible in two ways, corresponding to the 2 panels in the fourth row of fig. 6. And (3) designing 14 feature filters for enhancing watermark edge information according to the three structures, wherein the size of the feature filters is slightly larger than the watermark feature size, so that the features can be enhanced. Because 14 characteristic filters are designed, the characteristic filter pair is used for removing interferenceThe step of extracting and strengthening the feature map to obtain a response image comprises the following steps: extracting and strengthening the feature graph after removing the interference by using a feature filter, and obtaining a structural mode response image R of the feature graph to the feature filter for each feature filter i i Where i is 1,2, …,14, i.e. we finally obtain 14 corresponding response images R i 。
Further, the feature filter is weighted according to the center distance using a gaussian distribution. By the design, the watermark edge is more prominent, and the enhancement of the invisible watermark characteristic is realized. The effect of visualization using the gaussian distribution weighted according to the center distance is shown in fig. 6, and the gradual effect can be clearly seen.
Further, the step of calculating the watermark likelihood score S according to the response image includes the steps of: 14 response images R i The comprehensive response image R is obtained by combining the two images, and as shown in fig. 7, the value R (x, y) of the comprehensive response image R at the coordinate (x, y) is calculated according to the following formula: r (x, y) max [ R ] i (x,y)]In the formula, R i (x, y) is the value of the ith response image at the coordinate (x, y); averaging values of all coordinate positions of the comprehensive response image R to obtain a watermark possibility score S; when 14 response images R are to be displayed i When the single value S is obtained through the above steps, it can be conveniently determined, where the threshold T is a preset constant, and T is greater than or equal to 6 and less than or equal to 7, and more preferably, T is greater than or equal to 6.5. In practical application, a plurality of screen shot images (part of which is added with the invisible watermark and part of which is not added with the invisible watermark) can be used, the watermark possibility scores S corresponding to the images are obtained by processing according to the steps, and then the appropriate threshold value T is determined according to the scores and whether the watermarks exist, so that the value of T is more accurate.
Further, the invention also discloses a computer readable storage medium and an electronic device. A computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the feature map interference removal method or the screen shot watermark identification method as described above. An electronic device comprises a memory, a processor and a computer program stored on the memory, wherein the processor implements the feature pattern interference removal method or the screen shot watermark identification method when executing the computer program.
Claims (10)
1. A method for removing feature map interference is characterized by comprising the following steps: the method comprises the following steps:
carrying out self-adaptive binarization processing on an original image of the feature map to obtain a foreground map and taking the foreground map as a mask image;
according to the mask image, carrying out etching operation from outside to inside step by step with a fixed step length;
comparing the result after the etching operation with the mask image of the previous stage to obtain an interpolation region;
calculating the mean value filtering of the characteristic diagram interpolation area and filling the calculation result into the interpolation area;
and continuing the steps until the mask image is corroded to disappear, and obtaining the characteristic diagram after the interference is removed after multiple times of filling.
2. The signature interference removal method of claim 1, wherein: and in the step of carrying out self-adaptive binarization processing on the characteristic image to obtain a foreground image and taking the foreground image as a mask image, taking an image obtained by carrying out expansion processing on the foreground image as the mask image.
3. The signature interference removal method of claim 1, wherein: the fixed step length in the etching operation is 1-4 pixels.
4. The signature interference removal method of claim 1, wherein: the method comprises the following steps of carrying out self-adaptive binarization processing on an original image of a feature map to obtain a foreground map, and taking the foreground map as a mask image, wherein the feature map and the original image are obtained according to the following steps:
clipping and/or zooming the shot screen image to obtain an original image;
and (4) carrying out watermark extraction on the original image by using a watermark extraction program to obtain a characteristic image.
5. The signature interference removal method of claim 4, wherein: the shot screen image comprises characters, and a foreground image is a character area; the size of the original image is the same as the output size of the watermark extraction program.
6. A screen shot watermark identification method is characterized in that: the method comprises the following steps:
clipping and/or zooming the shot screen image to obtain an original image;
extracting the watermark of the original image by using a watermark extraction program to obtain a characteristic image;
processing the original image and the feature map according to the steps of claim 1 to obtain a feature map with interference removed;
constructing a characteristic filter corresponding to the characteristic size of the watermark;
extracting and strengthening the feature map after the interference is removed by using a feature filter to obtain a response image;
and calculating a watermark possibility score S according to the response image, comparing the S with a set threshold value T, judging that the screen shot image to be detected contains the watermark if the S is larger than the T, and otherwise, judging that the screen shot image to be detected does not contain the watermark.
7. The screen shot watermark recognition method of claim 6, wherein: the watermark in the screen shot image to be detected is block-shaped; the characteristic filter corresponding to the characteristic size of the watermark is constructed by the following steps:
determining possible structures according to the block-shaped watermarks, wherein the possible structures comprise corner structures, straight edge structures and chessboard-like structures;
designing 14 feature filters for reinforcing watermark edge information according to the three structures, wherein the size of each feature filter is larger than the feature size of the watermark;
the step of extracting and enhancing the feature map after the interference removal by using the feature filter to obtain a response image comprises the following steps: extracting and strengthening the feature graph after removing the interference by using a feature filter, and obtaining a structural mode response image R of the feature graph to the feature filter for each feature filter i i Where i is 1,2, …, 14.
8. The screen shot watermark recognition method of claim 7, wherein: the step of calculating the watermark possibility score S according to the response image comprises the following steps:
14 response images R i And (3) obtaining a comprehensive response image R by combining the images into a whole, wherein the value R (x, y) of the comprehensive response image R at the coordinate (x, y) is obtained by calculating according to the following formula: r (x, y) max [ R ] i (x,y)]In the formula, R i (x, y) is the value of the ith response image at the coordinate (x, y);
averaging values of all coordinate positions of the comprehensive response image R to obtain a watermark possibility score S;
the threshold value T is a constant, and T is more than or equal to 6 and less than or equal to 7.
9. A computer-readable storage medium characterized by: stored thereon a computer program which, when being executed by a processor, implements the signature interference removal method as claimed in any one of claims 1 to 5 or the screen shot watermark identification method as claimed in any one of claims 6 to 8.
10. An electronic device, characterized in that: comprising a memory, a processor and a computer program stored on the memory, the processor implementing the method for signature interference removal according to any of claims 1-5 or the method for screen shot watermark recognition according to any of claims 6-8 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210963058.0A CN115049837B (en) | 2022-08-11 | 2022-08-11 | Characteristic diagram interference removing method and screen shot watermark identification method comprising same |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210963058.0A CN115049837B (en) | 2022-08-11 | 2022-08-11 | Characteristic diagram interference removing method and screen shot watermark identification method comprising same |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115049837A true CN115049837A (en) | 2022-09-13 |
| CN115049837B CN115049837B (en) | 2022-11-18 |
Family
ID=83166393
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210963058.0A Active CN115049837B (en) | 2022-08-11 | 2022-08-11 | Characteristic diagram interference removing method and screen shot watermark identification method comprising same |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115049837B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1831818A (en) * | 2005-03-08 | 2006-09-13 | 富士施乐株式会社 | Translated document image production device, recording medium and translated document image production method |
| US20070223782A1 (en) * | 2002-07-26 | 2007-09-27 | Wataru Asano | Digital watermark detection method and apparatus |
| CN103186889A (en) * | 2011-12-30 | 2013-07-03 | Ge医疗系统环球技术有限公司 | Method and device for reducing metal artifacts in medical images |
| US20180131837A1 (en) * | 2014-11-19 | 2018-05-10 | Digimarc Corporation | Optimizing optical scanners for digital watermark detection |
| CN112017200A (en) * | 2020-08-07 | 2020-12-01 | 湖北省农业科学院农产品加工与核农技术研究所 | Fish posture judgment method based on visual recognition |
| CN112102141A (en) * | 2020-09-24 | 2020-12-18 | 腾讯科技（深圳）有限公司 | Watermark detection method, watermark detection device, storage medium and electronic equipment |
| CN114419006A (en) * | 2022-01-21 | 2022-04-29 | 四川九洲空管科技有限责任公司 | Method and system for removing watermark of gray level video characters changing along with background |
-
2022
- 2022-08-11 CN CN202210963058.0A patent/CN115049837B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070223782A1 (en) * | 2002-07-26 | 2007-09-27 | Wataru Asano | Digital watermark detection method and apparatus |
| CN1831818A (en) * | 2005-03-08 | 2006-09-13 | 富士施乐株式会社 | Translated document image production device, recording medium and translated document image production method |
| CN103186889A (en) * | 2011-12-30 | 2013-07-03 | Ge医疗系统环球技术有限公司 | Method and device for reducing metal artifacts in medical images |
| US20180131837A1 (en) * | 2014-11-19 | 2018-05-10 | Digimarc Corporation | Optimizing optical scanners for digital watermark detection |
| CN112017200A (en) * | 2020-08-07 | 2020-12-01 | 湖北省农业科学院农产品加工与核农技术研究所 | Fish posture judgment method based on visual recognition |
| CN112102141A (en) * | 2020-09-24 | 2020-12-18 | 腾讯科技（深圳）有限公司 | Watermark detection method, watermark detection device, storage medium and electronic equipment |
| CN114419006A (en) * | 2022-01-21 | 2022-04-29 | 四川九洲空管科技有限责任公司 | Method and system for removing watermark of gray level video characters changing along with background |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115049837B (en) | 2022-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI726422B (en) | Two-dimensional code recognition method, device and equipment | |
| CN108280829A (en) | Welding seam method, computer installation and computer readable storage medium | |
| CN110136069B (en) | Text image correction method and device and electronic equipment | |
| CN107545223B (en) | Image recognition method and electronic equipment | |
| CN108665495B (en) | Image processing method and device and mobile terminal | |
| CN111539238A (en) | Two-dimensional code image restoration method and device, computer equipment and storage medium | |
| WO2007047665A1 (en) | Modeling micro-structure for feature extraction | |
| CN112651953A (en) | Image similarity calculation method and device, computer equipment and storage medium | |
| CN112101323B (en) | Method, system, electronic device and storage medium for identifying title list | |
| WO2017128646A1 (en) | Image processing method and device | |
| CN115049837B (en) | Characteristic diagram interference removing method and screen shot watermark identification method comprising same | |
| CN115410191B (en) | Text image recognition method, device, equipment and storage medium | |
| CN111695550A (en) | Character extraction method, image processing device and computer readable storage medium | |
| CN116434071A (en) | Determination method, determination device, equipment and medium for normalized building mask | |
| CN116631003A (en) | Equipment identification method and device based on P & ID drawing, storage medium and electronic equipment | |
| CN115527023A (en) | Image detection method, image detection device, electronic equipment and storage medium | |
| CN115049840B (en) | Screen-shot watermark identification method, storage medium and electronic device | |
| CN111079756B (en) | Form extraction and reconstruction method and equipment in receipt image | |
| CN113971738A (en) | Image detection method, image detection device, electronic equipment and storage medium | |
| CN114648751A (en) | Method, device, terminal and storage medium for processing video subtitles | |
| CN112118478B (en) | Text processing method and device, electronic equipment and storage medium | |
| CN115908184B (en) | Automatic removal method and device for mole pattern | |
| CN112991451B (en) | Image recognition method, related device and computer program product | |
| CN113570703B (en) | Three-dimensional digital model porosity calculation method based on image recognition technology | |
| CN112001202B (en) | Network structure setting method for secondary face detection |
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 |