Intelligent monitoring method and device for power equipment

Abstract

The invention discloses an intelligent monitoring method and device for power equipment, and relates to the power monitoring technology, and the method comprises the steps: S1, collecting a video image of a power equipment site; s2, performing similarity analysis on the video image and a pre-stored image in a historical database, and if the similarity is greater than or equal to a set similarity threshold, taking a diagnosis result and a solution of the pre-stored image as a diagnosis result and a solution of the current video image; s3, carrying out blurred screen detection on the video image, and if the video image does not have a blurred screen, entering S4; s4, performing image still detection on the video image, and if the video image is a non-still image, entering S5; s5, performing anomaly detection on the video image, and if there is an anomaly, entering S6; and S6, configuring a diagnosis result and a solution for the video image, and carrying out abnormity positioning on the video image with a blurred screen or a static or abnormal image. According to the invention, the reliability and timeliness of monitoring are ensured, and the risk that the power equipment is abnormal on site is greatly reduced.

Description

Technical Field

[0001] The present invention relates to electric power monitoring technology, and more specifically, to an intelligent monitoring method and device for electric power equipment. Background Art

[0002] As the scale of power grids continues to expand, the reliability requirements for power equipment are also getting higher and higher. Due to the large number of power equipment and the large area of ​​the site, the traditional manual on-site supervision and inspection and post-event review management and control methods can hardly meet the real-time monitoring requirements of power equipment. Instead, the supervision and management requirements of power equipment on site are more timely, efficient and intelligent. Among them, video security monitoring can realize real-time monitoring and centralized management. It is one of the important ways to ensure the life safety of staff. It can realize timely alarm of safety incidents that endanger people and equipment, such as accidental entry into live areas, external hidden dangers at power equipment sites, and power equipment failures. Although the current video surveillance technology has been widely used in the power industry, it still remains at the stage of relying on traditional manual video monitoring. Its disadvantages are: monitoring requires special monitoring personnel to monitor 24 hours a day, which greatly wastes human resources; because there are many monitoring screens, a large amount of information, and most of the video data is normal data, the monitoring personnel rely on human eye monitoring and recognition. The efficiency is not high, and it is easy to miss key information, resulting in the problem that emergency situations cannot be reported and handled in time. Summary of the invention

[0003] In view of the deficiencies in the prior art, the present invention provides an intelligent monitoring method and device for electric power equipment, which ensures the reliability and timeliness of monitoring and greatly reduces the risk of on-site abnormalities of electric power equipment.

[0004] The present invention provides an intelligent monitoring method for electric power equipment, the method comprising the following steps:

[0005] S1, collecting video images of the power equipment site and preprocessing the video images;

[0006] S2, performing similarity analysis on the pre-processed video image and the pre-stored image in the history database, if the similarity is greater than or equal to the set similarity threshold, the diagnosis result and solution of the pre-stored image are used as the diagnosis result and solution of the current video image; otherwise, proceed to S3;

[0007] S3, performing a screen distortion detection on the pre-processed video image, if the screen distortion detection result is that the video image is not screen distortion, proceed to S4; otherwise, proceed to S6;

[0008] S4, performing image stillness detection on the pre-processed video image, if the image stillness detection result is that the video image is not still, proceed to S5; otherwise, proceed to S6;

[0009] S5, performing abnormality detection on the pre-processed video image, if there is an abnormality, proceed to S6;

[0010] S6. According to the distorted screen, still image or abnormality, the video image is configured with a diagnosis result and a solution, and then the pre-processed video image and the corresponding diagnosis result and solution are stored in a historical database, and the abnormality of the video image with the abnormality is located.

[0011] Preferably, the abnormality detection is specifically:

[0012] The video image is compared with a set standard image. If the comparison result shows that there is a foreign object in the video image, the part of the video image that is different from the standard image is extracted as a foreign object block, and the foreign object block is compared and analyzed with a pre-stored foreign object image block to identify the type of foreign object in the video image. Then, an abnormality analysis of the motion trajectory of the moving object on the scene or an abnormality analysis of the smoke on the scene is performed according to the type of the foreign object.

[0013] An infrared image based on the video image is acquired, and temperature anomaly analysis of the electric power equipment is performed according to the infrared image.

[0014] Preferably, if the foreign object type is a moving object, an abnormality analysis of the motion trajectory of the moving object on site is performed, specifically comprising the following steps:

[0015] S51, extracting multiple frames of video images with a time interval;

[0016] S52, extracting the positions of all moving objects in the video images, and depicting the actual motion trajectory of the moving objects according to the positions of the moving objects;

[0017] S53, performing motion trend simulation on the actual motion trajectory to extend it, and obtaining a predicted motion trajectory;

[0018] S54, identifying the object type of the moving object and constructing a safety area;

[0019] S55, determining whether the predicted motion trajectory interferes with an area outside the safe area, and if so, determining that the moving object is in abnormal motion, and executing an alarm mechanism.

[0020] Preferably, in S54, an image of the moving object is extracted, and the image is compared with object images in an object image database to identify the object type of the moving object, and an object type risk assessment value of the current video image is set according to the object type; a safe area range value of the moving object is obtained by calculating the product of the width of the movable area of ​​the moving object and the object type risk assessment value, and a safe area is constructed based on the safe area range value with the centerline position of the movable area at the power equipment site as a reference.

[0021] Preferably, if the foreign body type is smoke, the on-site smoke abnormality analysis is performed, which is specifically:

[0022] Set a scene transparency threshold, obtain the smoke transparency of the video image, and determine whether it is equal to or greater than the scene transparency threshold; if so, determine that the scene smoke is abnormal, and at the same time obtain the position with the minimum smoke transparency in the video image; if the distance between the position with the minimum smoke transparency and the power equipment is less than a set distance threshold, it is identified as abnormal smoke production by the power equipment, and an alarm mechanism is executed.

[0023] Preferably, the power equipment performs temperature anomaly analysis specifically as follows:

[0024] Acquire an infrared image corresponding to the video image; extract the image area of ​​the power equipment in the video image, and calibrate a target area consistent with the image area of ​​the power equipment in the infrared image based on the image area of ​​the power equipment; compare the target area with a preset standard infrared image, and if the color depth at the same position in the target area is greater than that of the standard infrared image, it is determined that the power equipment is heating abnormally, and an alarm mechanism is executed.

[0025] Preferably, the image stillness detection is specifically:

[0026] Acquire multiple video images with a time difference from the video image of the current frame as subsequent video sequences, convert all video images into grayscale images, and use the grayscale image of the video image of the current frame as a template; perform difference operations on the template and the grayscale images corresponding to the video images in the subsequent video sequences in turn, and count the mean and variance of all difference values ​​after each difference operation, and judge whether the mean and variance are less than the set corresponding threshold value; if the mean and variance are less than the set corresponding threshold value, count the number of continuous frames, otherwise judge the current video image to be in a non-static state; judge whether the number of continuous frames is equal to the set freeze frame requirement; if the number of continuous frames is equal to the set freeze frame requirement, issue a video image static alarm.

[0027] Preferably, in S6, the abnormality positioning is:

[0028] The text information of the camera corresponding to the video image with an abnormality is obtained, and the camera is checked according to the text information. If the check passes, the location information of the camera is searched from a database according to the text information.

[0029] Preferably, the text information includes name information and time information detection; the method of checking the camera is specifically:

[0030] The name information is matched with the name information preset in the system; if the match is successful, it is determined that the name information corresponds to the camera correctly, and then the time information is matched; if the match is wrong, it is determined that the name information corresponds to the camera incorrectly, and an alarm message is generated; the time information is matched with the actual time; if the match is successful, it is determined that the time information is correct, and then the location information of the camera is searched from the database according to the noun information, and the text information and the location information are sent to the server at the same time; if the match is unsuccessful, it is determined that the time information is wrong, and an alarm message is generated.

[0031] An intelligent monitoring device for electric power equipment comprises a plurality of cameras arranged at the electric power equipment site, a monitoring station communicating with the cameras, and a video monitoring center communicating with the monitoring station; the plurality of cameras send the collected video images and infrared images to the video monitoring through the monitoring station, and the video monitoring center applies the intelligent monitoring method to realize the monitoring of the electric power equipment and its site.

[0032] Beneficial Effects

[0033] The advantages of the present invention are: first, the collected video images are pre-processed and then a preliminary judgment is made through a similarity comparison. If there are similar video images in the historical database, the judgment result is directly output to obtain the diagnosis result and solution, and the solution is displayed, thereby achieving the effect of rapid monitoring and processing. In addition, for video images that do not pass the similarity, the monitoring reliability detection of power equipment and its on-site monitoring, as well as abnormality identification and abnormality location are realized in sequence through the flower screen, video stillness, and abnormality detection, so as to quickly obtain abnormal information and abnormal location, which is conducive to timely reminding video monitoring personnel and timely processing by on-site staff, ensuring the reliability and timeliness of monitoring, and greatly reducing the risk of power equipment causing on-site abnormalities. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the process flow of the intelligent monitoring method of power equipment of the present invention;

[0035] Figure 2 It is a schematic diagram of the image still detection process of the present invention;

[0036] Figure 3A schematic diagram of a safety zone constructed for the present invention;

[0037] Figure 4 It is a schematic diagram of the structure of the intelligent monitoring device for electric power equipment of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1 The present invention provides an intelligent monitoring method for electric power equipment, which includes the following steps.

[0040] S1. Collect video images of the power equipment site and perform filtering and denoising preprocessing on the video images.

[0041] S2, the pre-processed video image and the pre-stored image in the historical database are analyzed for similarity. If the similarity is greater than or equal to the set similarity threshold, such as when the similarity is greater than or equal to the set 80%, the diagnostic result and solution of the pre-stored image are used as the diagnostic result and solution of the current video image, which can greatly improve the processing speed of abnormal power equipment and achieve the effect of rapid monitoring and processing. If the similarity comparison fails, enter S3.

[0042] S3, performing a distorted screen detection on the pre-processed video image. If the distorted screen detection result is that the video image is not distorted, then proceed to S4.

[0043] The main reason for the abnormal video screen is noise interference. Video noise interference mainly detects the superimposed noise of the video image caused by factors such as line aging, transmission failure, poor contact, electromagnetic interference, etc., or the screen noise phenomenon such as snow, strip, and band-shaped stripe interference. Video noise interference occurs frequently and greatly affects the video quality. It is easy to cause other abnormal judgments of the video image. Therefore, it is necessary to detect the screen noise first to ensure that the monitoring equipment is working properly and the video image has no screen noise problem.

[0044] Considering that noise interference is accompanied by regular motion, we first extract the moving target of the image:

[0045] At any time t, calculate the difference image ΔI between the current time t and the time t-Δt t :

[0046] ΔI t =|It -I t-Δt |;

[0047] It represents the motion difference between the two images. We usually do the difference between consecutive video frames, the so-called inter-frame difference method to obtain the moving target.

[0048] Secondly, since the movement of noise is often high-frequency, we calculate the span average rate of pixels according to the formula:

[0049]

[0050] if[I t (x,y)-ave]*[I t+1 (x, y)-ave]<0, C t (x, y)+1;

[0051]

[0052] It(x,y) represents the pixel value of the grayscale image at (x,y) at time t, t1 represents the starting frame number, Δt represents the continuous frame number, and αve represents the pixel mean of the grayscale image at (x,y) for consecutive Δt frames.

[0053] According to the formula, the average number of spans of pixel I(x,y) in consecutive Δt frames is counted, and the average span rate p(x,y) of consecutive Δt frames is calculated according to the formula. Since noise belongs to high-frequency motion, when the average span rate of pixel I(x,y) exceeds the threshold T, it means that pixel I(x,y) is a high-frequency noise point.

[0054] S4, performing image stillness detection on the pre-processed video image. If the image stillness detection result is that the video image is not still, proceed to S5.

[0055] The video image remains still and constant, which we call a freeze frame. Once a freeze frame occurs, the surveillance video has lost its original surveillance effectiveness. Video transmission, camera failure or human intervention may cause the image to freeze.

[0056] For the video freeze frame phenomenon, we assume that a freeze frame occurs at time t, then the image I at time t t and the image I at time t+Δt t+Δt Exactly the same, including the time information in the upper left corner of the video image and the noise information in the image.

[0057] like Figure 2 As shown, in this embodiment, the specific method of image stillness detection is:

[0058] Acquire multiple video images with a time difference from the video image of the current frame as the subsequent video sequence, convert all video images into grayscale images, and use the grayscale image of the video image of the current frame as a template. Perform difference operations on the template and the grayscale images corresponding to the video images in the subsequent video sequence in turn, and count the mean and variance of all differences after each difference operation, and determine whether the mean and variance are less than the set corresponding threshold. If the mean and variance are less than the set corresponding threshold, count the number of continuous frames, otherwise determine that the current video image is not static; determine whether the number of continuous frames is equal to the set freeze frame requirement; if the number of continuous frames is equal to the set freeze frame requirement, issue a video image static alarm.

[0059] S5. Perform anomaly detection on the preprocessed video image. If anomaly exists, proceed to step S6.

[0060] In this embodiment, anomaly detection mainly includes three aspects, namely, moving objects, smoke, and the temperature of power equipment. The specific method of anomaly detection is:

[0061] The video image is compared with the set standard image. If the comparison result shows that there is a foreign object in the video image, the part of the video image that is different from the standard image is extracted as a foreign object block, and the foreign object block is compared and analyzed with the pre-stored foreign object image block to identify the type of foreign object in the video image. Then, according to the type of foreign object, an abnormal analysis of the motion trajectory of the moving object on site or an abnormal analysis of the smoke on site is performed. Since the monitoring of power equipment is mainly based on infrared images, this embodiment also needs to obtain an infrared image based on the video image, and perform a temperature abnormality analysis on the power equipment based on the infrared image. That is, when there is a foreign object in the video image, it is necessary to analyze the foreign object to determine whether it will cause abnormal behavior; and during the monitoring process, it is also necessary to monitor the temperature of the power equipment in real time to avoid safety hazards caused by abnormal temperature of the power equipment.

[0062] If the foreign body is a moving object, an abnormal analysis of the moving object's motion trajectory is performed on site, which specifically includes the following steps:

[0063] S51, extracting multiple frames of video images with a time interval.

[0064] S52, extracting the positions of the moving objects in all the video images, and depicting the actual motion trajectory of the moving objects according to the positions of the moving objects.

[0065] S53, performing motion trend simulation on the actual motion trajectory to extend it and obtain a predicted motion trajectory.

[0066] S54, identifying the object type of the moving object and constructing a safe area. The method for constructing the safe area is as follows:

[0067] Extract the image of the moving object and compare it with the object image in the object image database to identify the object type of the moving object, and set the object type risk assessment value of the current video image according to the object type. Obtain the safe area range value of the moving object by multiplying the width of the movable area of ​​the moving object and the object type risk assessment value, and construct the safe area based on the safe area range value with the center line position of the movable area of ​​the power equipment site as the reference.

[0068] like Figure 3 As shown in the figure, there will be a maintenance passage between the power equipment, which is the movable area of ​​the moving objects. The moving objects on the scene include staff, outsiders and other animals that stray into the scene. Among these moving objects, the staff are most familiar with the scene, so we can set their object type risk assessment value to 80%, that is, the safety area constructed by the center line position of the movable area, the spacing between the edge positions on both sides is 80% of the width of the movable area, and 10% of the width on both sides is left as the danger identification area. It is mainly considered that if the staff is too close to the power equipment during normal inspections, there will be hidden safety risks. Therefore, it is necessary to delineate the danger identification area during normal inspections to play a monitoring and reminder role in time when they are close to the power equipment. In addition, because outsiders are not familiar with power equipment, the object type risk assessment value can be set to 60% to reduce their safety risks on the scene. For the identification of staff and outsiders, the faces collected by the video image can be compared and identified, or they can be distinguished by the color of the safety helmet. Since other animals are prone to running around, the hidden dangers to power equipment are greater, so the object type risk assessment value can be set to 30% accordingly. By reducing the scope of its safe area, early warning can be achieved. If the predicted movement trajectory shows that the animal is always moving within the safe area, it means that the animal's movement trajectory is stable and has a tendency to leave the power equipment site on its own. There is no need to issue an alarm, avoiding the problem of panic and running around when the staff go to the site.

[0069] S55. Determine whether there is interference between the predicted motion trajectory and the area outside the safe area. If so, it is determined that the moving object is moving abnormally, and an alarm mechanism is executed, that is, the abnormality is displayed on the display screen of the video surveillance center to remind the monitoring personnel, and step S6 is executed at the same time to locate the abnormal point.

[0070] If the foreign body type is smoke, conduct on-site smoke anomaly analysis. The specific method is as follows:

[0071] Set a scene transparency threshold, obtain the smoke transparency of the video image, and determine whether it is equal to or greater than the scene transparency threshold. If so, it is determined that the scene smoke is abnormal, and the position with the minimum smoke transparency in the video image is obtained; if the distance between the position with the minimum smoke transparency and the power equipment is less than a set distance threshold, it is identified as abnormal smoke production by the power equipment, and the alarm mechanism is executed.

[0072] The specific method of performing temperature anomaly analysis on power equipment in this embodiment is:

[0073] Acquire an infrared image corresponding to the video image; extract the image area of ​​the power equipment in the video image, and calibrate the target area consistent with it in the infrared image based on the image area of ​​the power equipment; compare the target area with a preset standard infrared image. If the color depth at the same position in the target area is greater than that of the standard infrared image, it is judged that the power equipment is heating abnormally and an alarm mechanism is executed.

[0074] Through smoke detection and temperature detection, the operation of power equipment can be effectively monitored in real time to reduce the risk of abnormal operation.

[0075] S6. Configure the video image with diagnostic results and solutions, and then store the preprocessed video image and the corresponding diagnostic results and solutions in a historical database. The configuration of the diagnostic results and solutions can be completed by manual configuration, or the corresponding configuration scheme can be queried from the historical database according to the abnormality type, and configured to the abnormal video image as the diagnostic result and solution. If it is manually configured, after obtaining the diagnostic results and solutions of the abnormal video image, the historical database can be improved by storing the abnormal information, which is beneficial to the similarity analysis of the video image in the subsequent power equipment monitoring, thereby improving the pass rate of the similarity analysis and achieving the purpose of improving the abnormal monitoring and processing efficiency. After obtaining the corresponding diagnostic results and solutions, it is necessary to locate the abnormality of the video image with a flowery screen or a still or abnormal image to determine the actual position corresponding to the video image.

[0076] Specifically, the abnormality positioning is: obtaining the text information of the camera corresponding to the video image with the abnormality, verifying the camera according to the text information, and if the verification is passed, searching the location information of the camera from the database according to the text information.

[0077] Among them, the text information includes name information and time information detection. The way to check the camera is: match the name information with the name information preset in the system; if the match is successful, it is determined that the name information corresponds to the camera correctly, and then the time information is matched; if the match is wrong, it is determined that the name information corresponds to the camera incorrectly, and an alarm message is generated; match the time information with the actual time; if the match is successful, it is determined that the time information is correct, and then the location information of the camera is searched from the database according to the noun information, and the text information and location information are sent to the server at the same time; if the match is unsuccessful, it is determined that the time information is wrong, and an alarm message is generated.

[0078] It can be seen from steps S3-S6 that in view of the frequent abnormalities in the monitoring process, we first identify the abnormalities of the monitoring equipment, such as screen distortion and image stillness, and then identify the abnormalities of the monitored power equipment and its on-site. In this way, the abnormalities can be identified and eliminated one by one from the source of monitoring, and when an abnormality occurs in any link, the abnormality location is started, which greatly improves the timeliness of the monitoring of the power equipment, and can ensure that the abnormality is displayed and fed back to the on-site staff in the first time, thereby improving the efficiency of abnormal monitoring and processing, and ensuring the safe and reliable operation of the power equipment.

[0079] like Figure 4 As shown, an intelligent monitoring device for electric power equipment includes multiple cameras arranged at the electric power equipment site, a monitoring station communicating with the cameras, and a video monitoring center communicating with the monitoring station; the multiple cameras send the collected video images and infrared images to the video monitoring through the monitoring station, and the video monitoring center applies the above-mentioned intelligent monitoring method to realize the monitoring of the electric power equipment and its site, thereby improving the efficiency of abnormal monitoring and processing, and ensuring the safe and reliable operation of the electric power equipment.

[0080] The above description is a detailed description of the preferred feasible embodiments of the present invention, but the embodiments are not intended to limit the scope of the patent application of the present invention. All equivalent changes or modified changes completed under the technical spirit suggested by the present invention should fall within the patent scope covered by the present invention.

Claims

1. An intelligent monitoring method for power equipment, characterized in that: The method comprises the following steps: S1, collecting video images of the power equipment site and preprocessing the video images; S2, performing similarity analysis on the pre-processed video image and the pre-stored image in the history database, if the similarity is greater than or equal to the set similarity threshold, the diagnosis result and solution of the pre-stored image are used as the diagnosis result and solution of the current video image; otherwise, proceed to S3; S3, performing a screen distortion detection on the pre-processed video image, if the screen distortion detection result is that the video image is not screen distortion, proceed to S4; otherwise, proceed to S6; S4, performing image stillness detection on the pre-processed video image, if the image stillness detection result is that the video image is not still, proceed to S5; otherwise, proceed to S6; S5, performing abnormality detection on the pre-processed video image, if there is an abnormality, proceed to S6; S6. According to the distorted screen or still or abnormal image, the video image is configured with a diagnosis result and a solution, and then the preprocessed video image and the corresponding diagnosis result and solution are stored in a historical database, and the abnormality of the video image with distorted screen or still or abnormal image is located.

2. The intelligent monitoring method for electric power equipment according to claim 1, characterized in that: The anomaly detection is specifically as follows: The video image is compared with a set standard image. If the comparison result shows that there is a foreign object in the video image, the part of the video image that is different from the standard image is extracted as a foreign object block, and the foreign object block is compared and analyzed with a pre-stored foreign object image block to identify the type of foreign object in the video image. Then, an abnormality analysis of the motion trajectory of the moving object on the scene or an abnormality analysis of the smoke on the scene is performed according to the type of the foreign object. An infrared image based on the video image is acquired, and temperature anomaly analysis of the electric power equipment is performed according to the infrared image.

3. The intelligent monitoring method for electric power equipment according to claim 2, characterized in that: If the foreign object type is a moving object, an abnormality analysis of the motion trajectory of the moving object on site is performed, specifically including the following steps: S51, extracting multiple frames of video images with a time interval; S52, extracting the positions of all moving objects in the video images, and depicting the actual motion trajectory of the moving objects according to the positions of the moving objects; S53, performing motion trend simulation on the actual motion trajectory to extend it, and obtaining a predicted motion trajectory; S54, identifying the object type of the moving object and constructing a safety area; S55, determining whether the predicted motion trajectory interferes with an area outside the safe area, and if so, determining that the moving object is in abnormal motion, and executing an alarm mechanism.

4. The intelligent monitoring method for electric power equipment according to claim 3, characterized in that: In S54, extracting the image of the moving object, and comparing the image with the object image in the object image database to identify the object type of the moving object, and setting the object type risk assessment value of the current video image according to the object type; The safety area range value of the mobile object is obtained by calculating the product between the width of the movable area of ​​the mobile object and the object type risk assessment value, and the safety area is constructed based on the safety area range value with the center line position of the movable area at the power equipment site as a reference.

5. The intelligent monitoring method for electric power equipment according to claim 1, characterized in that: If the foreign body type is smoke, the on-site smoke anomaly analysis is performed, which is specifically as follows: Set a scene transparency threshold, obtain the smoke transparency of the video image, and determine whether it is equal to or greater than the scene transparency threshold; if so, determine that the scene smoke is abnormal, and at the same time obtain the position with the minimum smoke transparency in the video image; if the distance between the position with the minimum smoke transparency and the power equipment is less than a set distance threshold, it is identified as abnormal smoke production by the power equipment, and an alarm mechanism is executed.

6. The intelligent monitoring method for electric power equipment according to claim 1, characterized in that: The temperature anomaly analysis of the power equipment is specifically performed as follows: Acquire an infrared image corresponding to the video image; extract the image area of ​​the power equipment in the video image, and calibrate a target area consistent with the image area of ​​the power equipment in the infrared image based on the image area of ​​the power equipment; compare the target area with a preset standard infrared image, and if the color depth at the same position in the target area is greater than that of the standard infrared image, it is determined that the power equipment is heating abnormally, and an alarm mechanism is executed.

7. The intelligent monitoring method for electric power equipment according to claim 1, characterized in that: The image stillness detection is specifically as follows: Acquire multiple video images with a time difference from the video image of the current frame as subsequent video sequences, convert all video images into grayscale images, and use the grayscale image of the video image of the current frame as a template; perform difference operations on the template and the grayscale images corresponding to the video images in the subsequent video sequences in turn, and count the mean and variance of all difference values ​​after each difference operation, and judge whether the mean and variance are less than the set corresponding threshold value; if the mean and variance are less than the set corresponding threshold value, count the number of continuous frames, otherwise judge the current video image to be in a non-static state; judge whether the number of continuous frames is equal to the set freeze frame requirement; if the number of continuous frames is equal to the set freeze frame requirement, issue a video image static alarm.

8. The intelligent monitoring method for electric power equipment according to claim 1, characterized in that: In S6, the abnormality is located as follows: The text information of the camera corresponding to the video image with an abnormality is obtained, and the camera is checked according to the text information. If the check passes, the location information of the camera is searched from a database according to the text information.

9. The intelligent monitoring method for electric power equipment according to claim 8, characterized in that: The text information includes name information and time information detection; the method of checking the camera is specifically: The name information is matched with the name information preset in the system; if the match is successful, it is determined that the name information corresponds to the camera correctly, and then the time information is matched; if the match is wrong, it is determined that the name information corresponds to the camera incorrectly, and an alarm message is generated; the time information is matched with the actual time; if the match is successful, it is determined that the time information is correct, and then the location information of the camera is searched from the database according to the noun information, and the text information and the location information are sent to the server at the same time; if the match is unsuccessful, it is determined that the time information is wrong, and an alarm message is generated.

10. An intelligent monitoring device for electric power equipment, characterized in that: It includes multiple cameras arranged at the power equipment site, a monitoring station communicating with the cameras, and a video monitoring center communicating with the monitoring station; the multiple cameras send the collected video images and infrared images to the video monitoring through the monitoring station, and the video monitoring center uses the intelligent monitoring method as described in any one of claims 1 to 9 to realize the monitoring of the power equipment and its site.

Images