Intelligent judgment method and device for specific actions of personnel based on video

Abstract

The present invention relates to an intelligent judgment method and device for specific actions of personnel based on video. The method obtains real-time video information of each power equipment in a substation, captures human targets in the video information, judges whether the human targets in the video information make a warning starting posture, and if so, records the warning video information of the human targets after the warning starting posture, extracts the still frame in the warning video information, identifies the warning posture of the human targets in the still frame, deciphers the warning device fault information corresponding to the warning posture, and outputs the warning device fault information. In the case of an abnormality in the substation and a failure of the voice equipment of the patrol personnel, the patrol personnel can transmit information to the monitoring platform through the human body posture, and the monitoring platform can automatically obtain the warning device fault information according to the human body posture of the patrol personnel, so that the patrol personnel can quickly contact the monitoring platform behind and report the status in the substation, thereby ensuring the safety of the substation equipment and personnel.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric power safety, and in particular relates to an intelligent judgment method and device for specific actions of personnel based on video. Background Art

[0002] Substation refers to the place where voltage and current are transformed, electric energy is received and distributed in the power system. The substation in the power plant is a step-up substation, which is used to step up the electric energy generated by the generator and feed it into the high-voltage power grid. The inspection of the substation is that the on-duty personnel observe the appearance of the equipment for abnormalities through regular inspections, such as whether the color changes, whether there are debris, whether the needle indication is normal, whether the sound of the equipment is normal, whether there is abnormal smell, whether the temperature of the equipment allowed to be touched is normal, and measure the changes in the operating parameters of the electrical equipment during operation, etc., to determine whether the equipment is operating normally. The inspection system of the substation is an effective measure to ensure the normal and safe operation of the equipment. It is of great significance to reduce the occurrence of accidents and their impact range to understand the equipment operation status, grasp the abnormal operation, and take corresponding measures in time through regular inspections by the on-duty personnel. At present, the daily safety monitoring of substations is usually carried out by on-site inspections by inspection personnel and remote monitoring by the rear monitoring platform. The inspection personnel are equipped with voice equipment to communicate with the monitoring platform, so as to communicate with the on-duty personnel of the monitoring platform. However, the existing monitoring methods still have some blind spots. For example, when an abnormality occurs in the substation and the patrol personnel's voice equipment fails, they cannot quickly contact the rear monitoring platform and report the status of the substation. Therefore, how to overcome the shortcomings of the existing technology is an urgent problem to be solved in the field of power safety technology. Summary of the invention

[0003] The purpose of the present invention is to provide a method and device for intelligently judging specific actions of personnel based on video, so as to solve the problem that when an abnormality occurs in the substation and the voice equipment of the inspection personnel fails, they cannot quickly contact the rear monitoring platform and report the status in the substation.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] The intelligent judgment method of specific actions of personnel based on video includes:

[0006] Obtain real-time video information of each power equipment in the substation;

[0007] capturing a human target in the video information;

[0008] Determining whether a human target in the video information makes a warning starting gesture;

[0009] If the human target in the video information makes a warning starting gesture, record the warning video information of the human target after the warning starting gesture, and extract the still frame in the warning video information;

[0010] Identifying the warning posture of the human target in the still frame;

[0011] Deciphering the early warning device failure information corresponding to the early warning posture;

[0012] Output the early warning device failure information.

[0013] Further, preferably, after capturing the human target in the video information, the method further comprises:

[0014] Extracting facial information of the human target;

[0015] Based on the preset inspection personnel face database, the face information of the human target is recognized to determine whether the human target is an internal inspection personnel of the substation;

[0016] If the human target is an internal inspection personnel of the substation, it is determined whether the human target in the video information makes a warning starting gesture.

[0017] Further, preferably, identifying the warning posture of the human target in the still frame includes:

[0018] Extracting key skeleton points of the human body target in the still frame;

[0019] The key skeletal points are connected according to the human body structure to obtain a simplified diagram of the human body target posture.

[0020] Further, preferably, deciphering the warning device failure information corresponding to the warning posture includes:

[0021] Comparing the simplified human target posture image with a preset warning posture image database;

[0022] Determine a warning posture picture that matches the simplified posture picture of the human target;

[0023] Extract the warning device failure information corresponding to the warning posture image.

[0024] Further, preferably, there are multiple simplified diagrams of human target postures, and a warning posture picture corresponding to each simplified diagram of human target postures is used to represent a kind of warning device failure information, and the warning device failure information includes a device model and a device failure type.

[0025] The present invention also provides an intelligent judgment device for specific actions of personnel based on video, including:

[0026] An acquisition unit, used to acquire real-time video information of each power equipment in the substation;

[0027] A capturing unit, used for capturing a human target in the video information;

[0028] A first judgment unit, used to judge whether the human target in the video information makes a warning starting gesture;

[0029] A first extraction unit is used for recording the warning video information of the human target after the warning starting gesture in the video information and extracting a still frame in the warning video information;

[0030] A recognition unit, used for recognizing the warning posture of the human target in the still frame;

[0031] A cracking unit, used for cracking the early warning device failure information corresponding to the early warning posture;

[0032] An output unit is used to output the early warning device failure information.

[0033] Furthermore, preferably, the device further comprises:

[0034] A second extraction unit is used to extract face information of the human target after capturing the human target in the video information;

[0035] The second judgment unit is used to recognize the facial information of the human target based on a preset inspection personnel face database to determine whether the human target is an internal inspection personnel of the substation. In the case that the human target is an internal inspection personnel of the substation, the first judgment unit is used to determine whether the human target in the video information makes a warning starting gesture.

[0036] Further, preferably, the identification unit includes:

[0037] A third extraction unit is used to extract key skeleton points in the human target in the still frame;

[0038] The connection unit is used to connect the key skeleton points according to the human body structure to obtain a simplified diagram of the human body target posture.

[0039] Further, preferably, the cracking unit includes:

[0040] A comparison unit, used for comparing the simplified human target posture image with a preset warning posture image database;

[0041] A determination unit, used to determine a warning posture picture that matches the simplified human target posture picture;

[0042] The fourth extraction unit is used to extract the warning device failure information corresponding to the warning posture image.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention obtains real-time video information of each power equipment in the substation, captures human targets in the video information, determines whether the human targets in the video information make a warning starting gesture, and if the human targets in the video information make a warning starting gesture, records the warning video information of the human targets after the warning starting gesture, extracts the still frame in the warning video information, identifies the warning gesture of the human targets in the still frame, deciphers the warning device fault information corresponding to the warning gesture, and outputs the warning device fault information. When an abnormality occurs in the substation and the voice equipment of the inspectors fails, the inspectors can transmit information to the monitoring platform through the human body posture, and the rear monitoring platform can automatically obtain the warning device fault information according to the human body posture of the inspectors, so that the inspectors can quickly contact the rear monitoring platform and report the status in the substation, thereby ensuring the safety of the substation equipment and personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 It is a flow chart of the intelligent judgment method of personnel-specific actions based on video of the present invention;

[0047] Figure 2 A flowchart of an optional implementation of the method for intelligently determining specific actions of a person based on a video of the present invention;

[0048] Figure 3 It is a flow chart of another optional implementation of the method for intelligently judging specific actions of personnel based on video of the present invention;

[0049] Figure 4 A flowchart of another optional implementation of the method for intelligently determining specific actions of a person based on a video of the present invention;

[0050] Figure 5 Schematic diagram of the intelligent judgment device of personnel-specific actions based on video of the present invention. DETAILED DESCRIPTION

[0051] The present invention is further described in detail below in conjunction with embodiments.

[0052] Those skilled in the art will appreciate that the following examples are only used to illustrate the present invention and should not be considered to limit the scope of the present invention. If no specific techniques or conditions are specified in the examples, the techniques or conditions described in the literature in the art or the product specifications are used. If the manufacturer of the materials or equipment used is not specified, they are all conventional products that can be purchased.

[0053] See also Figure 1 The present invention provides an intelligent judgment method for specific actions of personnel based on video, the execution subject of the method is a monitoring platform server, and the method may include:

[0054] Step S101, obtaining real-time video information of each power equipment in the substation.

[0055] Specifically, a corresponding video image acquisition device can be set for each power equipment in the power station to collect real-time video information of the power equipment in real time. The video image acquisition device provides a network communication connection with the monitoring platform server and can transmit video information to the monitoring platform server.

[0056] Step S102, capturing a human target in the video information.

[0057] Specifically, after receiving the real-time video information of each power equipment in the substation sent by the video image acquisition device, the monitoring platform server captures the human targets in the video information so as to analyze the human targets later. The video image acquisition device may acquire infrared monitoring video, and specifically, a detection algorithm combining background subtraction and mathematical morphology may be used. Background subtraction is used to obtain the local area of ​​interest, and noise interference is further removed through morphology. False targets are eliminated through feature analysis, and the complete outline of the human body is extracted to obtain human target information. The background is updated using an adaptive background estimation model, thereby capturing the human targets in the video information. The specific detection algorithm can be implemented using existing technologies. Among them, background subtraction and mathematical morphology, feature analysis, and adaptive background estimation models are all existing technologies.

[0058] Specifically, Figure 2 As shown, after capturing the human target in the video information, the method further includes: step S201, extracting the face information of the human target. Step S202, based on the preset inspection personnel face database, the face information of the human target is recognized to determine whether the human target is an inspection personnel inside the substation. If the human target is an inspection personnel inside the substation, the subsequent step of determining whether the human target in the video information makes a warning starting gesture is executed. If the human target is not an inspection personnel inside the substation, the subsequent step of determining whether the human target in the video information makes a warning starting gesture does not need to be executed.

[0059] Step S103, determining whether the human target in the video information makes a warning starting gesture.

[0060] Specifically, the warning starting posture refers to the starting posture of the inspection personnel before making the warning posture. This posture is used to indicate that the warning posture is about to begin. The warning starting posture and various warning postures can be flexibly set and remembered by the inspection personnel. When needed, they can use their bodies to make corresponding postures in time to convey information.

[0061] Step S104: if the human target in the video information makes a warning start gesture, record the warning video information of the human target after the warning start gesture, and extract the still frame in the warning video information.

[0062] Specifically, if the human target in the video information makes a warning starting gesture, it means that various warning gestures will be made subsequently. Therefore, the warning video information of the human target after the warning starting gesture is recorded, and the still frames in the warning video information are extracted. When the inspection personnel make a certain warning gesture, they need to remain still for a few seconds to facilitate recognition by the monitoring platform server.

[0063] Step S105, identifying the warning posture of the human target in the still frame.

[0064] Specifically, Figure 3 As shown, the warning postures for identifying human targets in a still frame include:

[0065] Step S301, extracting key skeleton points in a human target in a still frame. The key skeleton points in the human target may include shoulders, elbows, head, trunk and knees, wherein the shoulder includes two skeleton key points, the elbow includes two skeleton key points, the head includes one skeleton key point, the trunk includes four skeleton key points, and the knee includes two skeleton key points.

[0066] Step S302, connecting key skeleton points according to the human body structure to obtain a simplified diagram of the human body target posture.

[0067] Step S106, deciphering the warning device failure information corresponding to the warning posture.

[0068] Specifically, Figure 4 As shown, the warning device failure information corresponding to the cracked warning posture includes:

[0069] Step S401, compare the simplified human body target posture image with a preset warning posture image database, wherein the preset warning posture image database stores various warning posture images.

[0070] Step S402, determining a warning posture image that matches the simplified human target posture image.

[0071] Step S403: extract the warning device failure information corresponding to the warning posture image, wherein each warning posture image corresponds to the text of the corresponding warning device failure information.

[0072] The simplified human target posture diagram may be multiple, and the warning posture image corresponding to each simplified human target posture diagram is used to represent a warning device failure information, and the warning device failure information includes the device model and the device failure type. For example, after the inspection personnel make the warning starting posture, they first make the first warning posture, which is used to convey the device model, and then make the second warning posture, which is used to convey the device failure type.

[0073] Step S107, outputting warning equipment failure information.

[0074] like Figure 5 As shown, the present invention provides an intelligent judgment device for specific actions of personnel based on video, comprising:

[0075] The acquisition unit 51 is used to acquire real-time video information of each power device in the substation.

[0076] The capturing unit 52 is used to capture the human target in the video information.

[0077] The first judging unit 53 is used to judge whether the human target in the video information makes a warning starting gesture.

[0078] The first extraction unit 54 is used for recording the warning video information of the human target after the warning starting gesture and extracting the still frame in the warning video information when the human target in the video information makes a warning starting gesture.

[0079] The recognition unit 55 is used to recognize the warning posture of the human target in the still frame.

[0080] The decryption unit 56 is used to decrypt the early warning device failure information corresponding to the early warning posture.

[0081] The output unit 57 is used to output warning device failure information.

[0082] Specifically, the device also includes:

[0083] The second extraction unit is used to extract the face information of the human target after capturing the human target in the video information.

[0084] The second judgment unit is used to recognize the facial information of the human target based on the preset inspection personnel face database to determine whether the human target is an internal inspection personnel of the substation. In the case that the human target is an internal inspection personnel of the substation, the first judgment unit is used to determine whether the human target in the video information makes a warning starting gesture.

[0085] Specifically, the identification unit includes:

[0086] The third extraction unit is used to extract key skeleton points in the human target in the still frame.

[0087] The connection unit is used to connect the key bone points according to the human body structure to obtain a simplified diagram of the human body target posture.

[0088] Specifically, the cracking unit includes:

[0089] The comparison unit is used to compare the simplified image of the human target posture with a preset warning posture image database.

[0090] The determination unit is used to determine a warning posture image that matches the simplified posture image of the human target.

[0091] The fourth extraction unit is used to extract the warning device fault information corresponding to the warning posture image.

[0092] The embodiment of the present invention also provides a storage medium, and the embodiment of the present invention also provides a storage medium, wherein a computer program is stored in the storage medium, and when the computer program is executed by a processor, some or all of the steps in each embodiment of the intelligent judgment method of specific actions of personnel based on video provided by the present invention are implemented. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0093] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution in the embodiments of the present invention can be essentially or partly contributed to the prior art in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0094] In this specification, the same or similar parts between the various embodiments can be referred to each other. In particular, for the embodiment of the intelligent judgment device for specific actions of personnel based on video, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method embodiment.

[0095] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An intelligent judgment method performed by an intelligent judgment device for specific actions of a person based on a video, characterized in that: The intelligent judgment device for specific actions of personnel in the video includes: An acquisition unit, used to acquire real-time video information of each power equipment in the substation; A capturing unit, used for capturing a human target in the video information; A first judgment unit, used to judge whether the human target in the video information makes a warning starting gesture; A first extraction unit is used for recording the warning video information of the human target after the warning starting gesture in the video information and extracting a still frame in the warning video information; A recognition unit, used for recognizing the warning posture of the human target in the still frame; A cracking unit, used for cracking the early warning device failure information corresponding to the early warning posture; An output unit, used to output the early warning device failure information; A second extraction unit is used to extract face information of the human target after capturing the human target in the video information; The capturing of human targets in the video information comprises: eliminating false targets through feature analysis, extracting the complete outline of the human body, and obtaining human target information; The second judgment unit is used to recognize the facial information of the human target based on a preset inspection personnel face database to determine whether the human target is an internal inspection personnel of the substation. If the human target is an internal inspection personnel of the substation, the first judgment unit is used to determine whether the human target in the video information makes a warning starting gesture; A third extraction unit is used to extract key skeleton points in the human target in the still frame; The key bone points in the human target include shoulders, elbows, head, trunk and knees, wherein the shoulder includes two bone key points, the elbow includes two bone key points, the head includes one bone key point, the trunk includes four bone key points, and the knee includes two bone key points; A connection unit, used to connect the key skeletal points according to the human body structure to obtain a simplified diagram of the human body target posture; The decryption unit comprises: A comparison unit, used for comparing the simplified human target posture image with a preset warning posture image database; A determination unit, used to determine a warning posture picture that matches the simplified human target posture picture; A fourth extraction unit, used to extract the warning device fault information corresponding to the warning posture picture; The intelligent judgment method comprises: Obtain real-time video information of each power equipment in the substation; capturing a human target in the video information; Determining whether a human target in the video information makes a warning starting gesture; If the human target in the video information makes a warning start gesture, record the warning video information of the human target after the warning start gesture, and extract the still frame in the warning video information; when the patrol personnel makes a certain warning gesture, they need to remain still for a few seconds to facilitate the recognition of the monitoring platform server; Identifying the warning posture of the human target in the still frame; Deciphering the early warning device failure information corresponding to the early warning posture; Outputting the early warning device failure information; in: The identifying of the warning posture of the human target in the still frame comprises: extracting key skeleton points of the human target in the still frame; connecting the key skeleton points according to the human body structure to obtain a simplified diagram of the human target posture; The deciphering of the warning device fault information corresponding to the warning posture comprises: comparing the simplified human target posture diagram with a preset warning posture picture database; determining a warning posture picture that matches the simplified human target posture diagram; Extracting the warning device fault information corresponding to the warning posture image; After capturing the human target in the video information, the method further includes: Extracting facial information of the human target; Based on the preset inspection personnel face database, the face information of the human target is recognized to determine whether the human target is an internal inspection personnel of the substation; If the human target is an internal inspection personnel of the substation, it is determined whether the human target in the video information makes a warning starting gesture; There are multiple simplified diagrams of human target postures, and the warning posture picture corresponding to each simplified diagram of human target postures is used to represent a kind of warning device failure information, and the warning device failure information includes a device model and a device failure type.

Images