CN113516009A - Method for eliminating influence of watermark on image of power transmission line - Google Patents
Method for eliminating influence of watermark on image of power transmission line Download PDFInfo
- Publication number
- CN113516009A CN113516009A CN202110374052.5A CN202110374052A CN113516009A CN 113516009 A CN113516009 A CN 113516009A CN 202110374052 A CN202110374052 A CN 202110374052A CN 113516009 A CN113516009 A CN 113516009A
- Authority
- CN
- China
- Prior art keywords
- watermark
- hidden danger
- transmission line
- image
- power transmission
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000005540 biological transmission Effects 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 19
- 238000001914 filtration Methods 0.000 claims description 5
- 230000007547 defect Effects 0.000 abstract description 3
- 238000009825 accumulation Methods 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000012790 confirmation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Landscapes
- Editing Of Facsimile Originals (AREA)
Abstract
The invention discloses a method for eliminating the influence of watermark on an image of a power transmission line, which screens hidden danger alarms possibly caused by the watermark based on the identified hidden danger types and hidden danger positions, confirms whether the hidden danger alarms caused by the watermark are hidden danger alarms or not by carrying out Tesseract-OCR character identification on the hidden danger alarms, and further eliminates the influence of the watermark on an identification result. Aiming at the technical defect that image watermarks are difficult to distinguish by adopting single hidden danger identification in the prior art, the invention can accurately distinguish the image watermarks from real wire foreign matters, avoid accumulation of wrong hidden danger information and improve the efficiency of artificial auxiliary identification.
Description
Technical Field
The invention discloses a method for eliminating the influence of watermark on an image of a power transmission line, and belongs to the technical field of detection of hidden danger of the power transmission line.
Background
With the continuous development of the technology for identifying the hidden danger of the channel of the power transmission line, the technology for automatically identifying and alarming the hidden danger is widely applied to the field of power transmission line detection, and the hidden danger which threatens a wire, such as a foreign wire, a crane and the like, can be accurately identified at present. In the actual image identification process, the watermark information on the image has great influence on the identification result, the part where the watermark is overlapped with the wire is easily identified as the wire foreign matter, and the manpower resource is wasted to confirm whether the hidden danger of the real wire foreign matter exists.
The existing solution is to increase positive and negative samples of the foreign matter on the wire to continue training the recognition model, so as to improve the recognition effect of the model, but the recognition model is difficult to distinguish for the foreign matter on the wire with the shape similar to the watermark.
In summary, how to provide an efficient and reliable method for eliminating watermark influence on a power transmission line channel to improve the reliability of a hidden danger identification result is one of the problems to be solved by the technical staff in the field at present.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method for eliminating the influence of the watermark on the image of the power transmission line. The method can eliminate the influence of the watermark on the identification result in the process of identifying the hidden danger of the power transmission line channel.
Summary of The Invention
The method screens the hidden danger alarms possibly caused by the watermark based on the identified hidden danger types and hidden danger positions, confirms whether the hidden danger alarms caused by the watermark are the hidden danger alarms caused by the watermark or not by carrying out Tesseract-OCR character identification on the hidden danger alarms, and further eliminates the influence of the watermark on the identification result.
The detailed technical scheme of the invention is as follows:
a method for eliminating the influence of watermark on the image of a power transmission line is characterized by comprising the following steps:
a. identifying the power transmission line channel image by using the existing hidden danger identification software aiming at the power transmission line channel image to obtain a hidden danger identification result;
b. extracting the hidden danger of the foreign matters of the wires in the identification result and giving an alarm;
c. judging whether a region corresponding to the wire foreign matter hidden danger alarm is superposed with a watermark position of a transmission line channel image, wherein the watermark position is information preset and stored in a system;
when the area is coincident with the watermark position, skipping to the step d;
when the area is not coincident with the watermark position, skipping to the step e;
d. c, carrying out character recognition on the hidden danger of the watermark position area, and filtering the hidden danger recognition result of the alarm corresponding to the recognized character and/or data symbol in the step a;
e. and (6) ending.
According to the invention, the character recognition is preferably Tesseract-OCR character recognition.
According to the present invention, the watermark position is the information preset to be stored in the system, including but not limited to: the coordinate range of the watermark position in the power transmission line channel image and the data rule of the display content of the watermark position, wherein the data rule comprises the type of character data, the arrangement rule of the character data and the data symbols among the character data.
According to the optimization of the invention, when the region is partially overlapped with the watermark position, the step d is skipped, and the watermark position is called to carry out information auxiliary to carry out character recognition on the hidden danger of the watermark position region.
Compared with the prior art, the invention has the following technical advantages:
aiming at the technical defect that image watermarks are difficult to distinguish by adopting single hidden danger identification in the prior art, the invention can accurately distinguish the image watermarks from real wire foreign matters, avoid accumulation of wrong hidden danger information and improve the efficiency of artificial auxiliary identification.
The invention introduces the concept of watermark position information in a breakthrough way, carries out range identification on the identified hidden danger information and the watermark position, and only the hidden danger information falling into the watermark position needs to be confirmed in a character identification way, if characters and/or data symbols exist, corresponding hidden dangers are directly filtered.
In addition, the invention also introduces a Tesseract-OCR character recognition algorithm, can quickly confirm whether the hidden danger is the watermark, and obviously improves the recognition speed compared with a hidden danger recognition model.
In order to further improve the identification effect of the invention, the invention further prefers to perform system prestoring on watermark position information, and when the region is partially overlapped with the watermark position, the watermark position can be called to perform information auxiliary to perform character identification on the hidden danger of the watermark position region.
Drawings
Fig. 1 is a schematic flow chart of a method for eliminating the influence of watermark on an image of a power transmission line according to the present invention;
FIG. 2 shows that in step a of the present invention, a hidden danger identification result is obtained by identifying a transmission line channel image using a conventional hidden danger identification model for the transmission line channel image;
in fig. 2, the collected image of a certain transmission line channel is shown, wherein the area at the upper left corner of the image is the watermark position, wherein "2019-10-14" is the collection date of the image, "13: 40: 13" is the collection time of the image, and the "frame" on the collection date in the image is the content filtered by the method of the present invention; the record "colorbeads 0.86" is a foreign matter of the wire, and the confidence rate is 0.86.
Detailed Description
The invention is further described below, but not limited thereto, with reference to the following examples and the accompanying drawings.
Examples 1,
As shown in fig. 1.
a. Identifying the power transmission line channel image by using the existing hidden danger identification software aiming at the power transmission line channel image to obtain a hidden danger identification result;
b. extracting the hidden danger of the foreign matters of the wires in the identification result and giving an alarm;
c. judging whether a region corresponding to the wire foreign matter hidden danger alarm is superposed with a watermark position of a transmission line channel image, wherein the watermark position is information preset and stored in a system;
when the area is coincident with the watermark position, skipping to the step d;
when the area is not coincident with the watermark position, skipping to the step e;
d. c, carrying out character recognition on the hidden danger of the watermark position area, and filtering the hidden danger recognition result of the alarm corresponding to the recognized character and/or data symbol in the step a;
e. and (6) ending.
The character recognition is Tesseract-OCR character recognition.
The watermark location is information preset to be stored in the system, including but not limited to: the coordinate range of the watermark position in the power transmission line channel image and the data rule of the display content of the watermark position, wherein the data rule comprises the type of character data, the arrangement rule of the character data and the data symbols among the character data. In the present embodiment, the watermark location information includes a number and "-".
Application example:
the method in the embodiment 1 is applied to images shot in one day in a certain power transmission line channel, 120 wire foreign matter alarms are totally obtained after hidden danger identification, the wire foreign matter alarms are extracted, position judgment and Tesseract-OCR character identification are carried out, the condition that the hidden danger alarms are caused by watermarks is filtered, and the influence of the watermarks on identification results is eliminated.
a. Extracting the hidden danger of the foreign matters of the wires in the identification result and giving an alarm;
the source program code of the alarm data is displayed as follows:
the image display result is shown in the attached figure 2;
the field meanings of the source program code are shown in the table 1:
table 1: field meaning of source program code
Field(s) | Type (B) | Means of |
filename | string | Filename |
width | int | Image width |
height | int | Image height |
objects.tag | string | Type of hidden danger |
objects.score | float | Confidence rate |
objects.frame.x | int | Coordinate x of upper left corner of alarm box |
objects.frame.y | int | Alarm box upper left corner coordinate y |
objects.frame.width | int | Alarm frame width |
objects.frame.height | int | Alarm frame height |
b. Judging whether the wire foreign matter hidden danger alarm is positioned at the watermark position on the upper left of the image, and screening the wire foreign matter hidden danger alarm in the watermark region; wherein, the watermark position area is: xmin: 30, ymin: 50, xmax: 600, ymax: 100, respectively;
c. performing Tesseract-OCR character recognition on the hidden danger of the watermark position, and filtering the alarm containing numbers or a' -symbol in the recognition result, wherein the Tesseract-OCR character recognition result is as follows: '2019-10-21' and '10-14';
d. and filtering the alarm corresponding to the number or the '-' symbol contained in the identification result.
In the application example, a Tesseract-OCR character recognition algorithm is adopted, so that the influence on the transmission line channel watermark caused by the result is eliminated. The problem of hidden danger warning caused by watermarks is well solved, the automation level of the warning detection of the power transmission line is improved, and the time cost for operation and maintenance personnel to confirm the warning is reduced.
Through statistics, in the application example, the false alarm rate of identifying the foreign matters of the model wires is reduced by 0.5 through eliminating the influence of the watermarks. The confirmation rate of the hidden danger alarm caused by the watermark reaches 0.89.
Examples 2,
The method of embodiment 1, further comprising: and when the region is overlapped with the watermark position part, skipping to the step d, calling the watermark position to carry out information auxiliary to carry out character recognition on the hidden danger of the watermark position region. For example: when the shape or the pattern of the foreign matter of the wire is recorded with characters and symbols, unnecessary interference is caused to the character recognition in the step d, therefore, the invention calls the watermark position information pre-stored in the system, and by verifying which contents are related to the character data and the symbol data specifically contained in the watermark position, the contents of the characters and the symbols are not filtered even if the characters are recognized, thereby improving the recognition effect.
Claims (4)
1. A method for eliminating the influence of watermark on the image of a power transmission line is characterized by comprising the following steps:
a. identifying the power transmission line channel image by using the existing hidden danger identification software aiming at the power transmission line channel image to obtain a hidden danger identification result;
b. extracting the hidden danger of the foreign matters of the wires in the identification result and giving an alarm;
c. judging whether a region corresponding to the wire foreign matter hidden danger alarm is superposed with a watermark position of a transmission line channel image, wherein the watermark position is information preset and stored in a system;
when the area is coincident with the watermark position, skipping to the step d;
when the area is not coincident with the watermark position, skipping to the step e;
d. c, carrying out character recognition on the hidden danger of the watermark position area, and filtering the hidden danger recognition result of the alarm corresponding to the recognized character and/or data symbol in the step a;
e. and (6) ending.
2. The method for eliminating the influence of the watermark on the image of the power transmission line according to claim 1, wherein the character recognition is Tesseract-OCR character recognition.
3. The method for eliminating the influence of the watermark on the image of the power transmission line according to claim 1 or 2, wherein the watermark position is information preset and stored in a system, and the information includes but is not limited to: the coordinate range of the watermark position in the power transmission line channel image and the data rule of the display content of the watermark position, wherein the data rule comprises the type of character data, the arrangement rule of the character data and the data symbols among the character data.
4. The method for eliminating the influence of the watermark on the image of the power transmission line according to claim 3, wherein when the area partially coincides with the watermark position, the step d is skipped, and the watermark position is called to perform information-assisted character recognition on the hidden danger of the watermark position area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110374052.5A CN113516009A (en) | 2021-04-07 | 2021-04-07 | Method for eliminating influence of watermark on image of power transmission line |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110374052.5A CN113516009A (en) | 2021-04-07 | 2021-04-07 | Method for eliminating influence of watermark on image of power transmission line |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113516009A true CN113516009A (en) | 2021-10-19 |
Family
ID=78062208
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110374052.5A Pending CN113516009A (en) | 2021-04-07 | 2021-04-07 | Method for eliminating influence of watermark on image of power transmission line |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113516009A (en) |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050018845A1 (en) * | 2003-07-01 | 2005-01-27 | Oki Electric Industry Co., Ltd. | Electronic watermark embedding device, electronic watermark detection device, electronic watermark embedding method, and electronic watermark detection method |
CN101246556A (en) * | 2008-01-21 | 2008-08-20 | 中山大学 | Image authentication system based on high-capacity USB KEY |
CN102495833A (en) * | 2011-11-30 | 2012-06-13 | 哈尔滨工业大学 | Document watermark copyright information protection device based on Opentype vector outline fonts |
CN103295199A (en) * | 2013-05-29 | 2013-09-11 | 西安建筑科技大学 | Intelligent repair assistance system for cracks of ancient wall murals |
CN103324873A (en) * | 2013-07-15 | 2013-09-25 | 上海北丞电子发展有限公司 | Copyright administration watermark publishing system and watermark embedding method |
CN106231195A (en) * | 2016-08-15 | 2016-12-14 | 乐视控股(北京)有限公司 | A kind for the treatment of method and apparatus of taking pictures of intelligent terminal |
CN106599780A (en) * | 2016-10-27 | 2017-04-26 | 国家电网公司 | Power grid polling image intelligent identification method and device |
CN107346528A (en) * | 2017-07-06 | 2017-11-14 | 安徽大学 | A kind of distorted image detection method based on the reversible fragile watermark of digraph |
CN109285105A (en) * | 2018-09-05 | 2019-01-29 | 北京字节跳动网络技术有限公司 | Method of detecting watermarks, device, computer equipment and storage medium |
CN112436600A (en) * | 2020-10-29 | 2021-03-02 | 山东理工大学 | Power equipment monitoring system based on flight inspection mode |
CN112446262A (en) * | 2019-09-02 | 2021-03-05 | 深圳中兴网信科技有限公司 | Text analysis method, text analysis device, text analysis terminal and computer-readable storage medium |
-
2021
- 2021-04-07 CN CN202110374052.5A patent/CN113516009A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050018845A1 (en) * | 2003-07-01 | 2005-01-27 | Oki Electric Industry Co., Ltd. | Electronic watermark embedding device, electronic watermark detection device, electronic watermark embedding method, and electronic watermark detection method |
CN101246556A (en) * | 2008-01-21 | 2008-08-20 | 中山大学 | Image authentication system based on high-capacity USB KEY |
CN102495833A (en) * | 2011-11-30 | 2012-06-13 | 哈尔滨工业大学 | Document watermark copyright information protection device based on Opentype vector outline fonts |
CN103295199A (en) * | 2013-05-29 | 2013-09-11 | 西安建筑科技大学 | Intelligent repair assistance system for cracks of ancient wall murals |
CN103324873A (en) * | 2013-07-15 | 2013-09-25 | 上海北丞电子发展有限公司 | Copyright administration watermark publishing system and watermark embedding method |
CN106231195A (en) * | 2016-08-15 | 2016-12-14 | 乐视控股(北京)有限公司 | A kind for the treatment of method and apparatus of taking pictures of intelligent terminal |
CN106599780A (en) * | 2016-10-27 | 2017-04-26 | 国家电网公司 | Power grid polling image intelligent identification method and device |
CN107346528A (en) * | 2017-07-06 | 2017-11-14 | 安徽大学 | A kind of distorted image detection method based on the reversible fragile watermark of digraph |
CN109285105A (en) * | 2018-09-05 | 2019-01-29 | 北京字节跳动网络技术有限公司 | Method of detecting watermarks, device, computer equipment and storage medium |
CN112446262A (en) * | 2019-09-02 | 2021-03-05 | 深圳中兴网信科技有限公司 | Text analysis method, text analysis device, text analysis terminal and computer-readable storage medium |
CN112436600A (en) * | 2020-10-29 | 2021-03-02 | 山东理工大学 | Power equipment monitoring system based on flight inspection mode |
Non-Patent Citations (1)
Title |
---|
段春梅;张涛川;李大成;: "基于时频域混合分析的太阳能硅片缺陷检测方法", 机床与液压, no. 08 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112434695B (en) | Upper pull rod fault detection method based on deep learning | |
CN111815605B (en) | Sleeper defect detection method based on step-by-step deep learning and storage medium | |
CN105654072A (en) | Automatic character extraction and recognition system and method for low-resolution medical bill image | |
CN107958201A (en) | A kind of intelligent checking system and method for vehicle annual test insurance policy form | |
CN111080620A (en) | Road disease detection method based on deep learning | |
CN112364740B (en) | Unmanned aerial vehicle room monitoring method and system based on computer vision | |
CN112733668B (en) | Video deep learning-based detection method for individual elevator taking of pets in elevator | |
CN110310255A (en) | Notch of switch machine detection method based on target detection and image procossing | |
CN107886047A (en) | A kind of detecting system and method for vehicle annual test survey report | |
CN102780856B (en) | Method for annotating subtitles in news video | |
CN113298077A (en) | Transformer substation foreign matter identification and positioning method and device based on deep learning | |
CN113516009A (en) | Method for eliminating influence of watermark on image of power transmission line | |
JP2006292593A (en) | Flaw detector, flaw detecting method, computer program and recording medium | |
CN104298993A (en) | Pole number positioning and identification method suitable for railway line in complex scene | |
CN115995056A (en) | Automatic bridge disease identification method based on deep learning | |
CN110378862A (en) | A kind of raising transmission line of electricity breaks the data enhancement methods of target identification accuracy outside | |
CN116383797B (en) | Non-notch sliding verification code and generation method thereof | |
CN111626104B (en) | Cable hidden trouble point detection method and device based on unmanned aerial vehicle infrared thermal image | |
CN112251861A (en) | Broken end detection device and detection method for ring spinning frame | |
CN112906562B (en) | Safety early warning method for side plate crack of trolley of chain grate | |
CN116664512A (en) | VEH connector assembly safety monitoring alarm system based on visual identification | |
CN112784914B (en) | Pipe gallery video intelligent attribute detection method and system based on cloud processing | |
CN112967224A (en) | Electronic circuit board detection system, method and medium based on artificial intelligence | |
CN112633115B (en) | Method and device for detecting vehicle bottom carrying object and storage medium | |
CN115689994B (en) | Nameplate and bar code defect detection method, equipment and storage medium |
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 |