Power plant unsafe behavior platform-based unsafe behavior detection method and device for flammable and explosive area, storage medium and processor

Abstract

The invention discloses an unsafe behavior detection method and device for a flammable and explosive area based on a power plant unsafe behavior platform, a storage medium and a processor, relates to the technical field of operation safety management, and mainly aims at improving the accuracy of unsafe behavior detection of the flammable and explosive area and effectively preventing accidents. According to the main technical scheme, effective image frames of a target range are obtained, the target range is an entrance management range of an inflammable and explosive area, and the effective image frames are used for representing continuous image frames of pedestrian targets selected by boundary frames; detecting whether the pedestrian target triggers a non-safe behavior corresponding to the target range or not according to the effective image frame, wherein the non-safe behavior is set according to the safe operation requirement of the flammable and explosive area; and if yes, executing a corresponding alarm operation based on the behavior type of the pedestrian target triggering non-safety behavior. The method is used for non-safe behavior detection of flammable and explosive areas based on the power plant non-safe behavior platform.

Description

Technical Field

[0001] The present application relates to the technical field of operation safety management, and in particular to a method, device, storage medium and processor for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant. Background Art

[0002] In the management of flammable and explosive areas, safety issues are always the primary concern, because once an accident occurs in such areas, it often leads to significant property losses and casualties. Traditional safety management relies on manual monitoring, especially in places such as gas-fired thermal power plants, where the operating environment is complex and there are high-risk areas where flammable and explosive gases are stored. Before entering these areas, operators must strictly abide by safety behavior standards, such as wearing safety helmets and work clothes correctly, thoroughly discharging static electricity, and storing personal items that may cause fire or explosion, such as mobile phones and lighters.

[0003] At present, the existing unsafe behavior detection in flammable and explosive areas based on the unsafe behavior platform of power plants usually relies on the behavior recognition of the monitoring system, that is, real-time capture of the corresponding video image frames, identification and judgment of whether the behavior in the image frame is unsafe behavior. However, this method is mostly based on the analysis of a single image frame, lacks consideration of the continuity of the action, and is difficult to accurately identify complex dynamic behaviors. The safety behavior specifications corresponding to flammable and explosive areas mostly need to last for a certain period of time. For example, the release of static electricity usually requires continuous contact with the static ball for more than 3 seconds, etc. This leads to low detection accuracy of unsafe behaviors in flammable and explosive areas, increasing the probability of accidents. Summary of the invention

[0004] In view of the above problems, the present application provides a method, device, storage medium and processor for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant. The main purpose is to improve the accuracy of unsafe behavior detection in flammable and explosive areas and effectively prevent accidents.

[0005] In order to solve the above technical problems, this application proposes the following solutions:

[0006] In a first aspect, the present application provides a method for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant, the method comprising:

[0007] Acquire valid image frames of a target range, where the target range is an entrance management range of an inflammable and explosive area, and the valid image frames are used to represent continuous image frames in which a pedestrian target is selected by a bounding box;

[0008] Detecting, according to the valid image frame, whether the pedestrian target triggers an unsafe behavior corresponding to the target range, wherein the unsafe behavior is set according to the safety operation requirements of the flammable and explosive area;

[0009] If so, a corresponding warning operation is performed based on the behavior type of the pedestrian target triggering the unsafe behavior.

[0010] In a second aspect, the present application provides a non-safe behavior detection device for flammable and explosive areas based on an unsafe behavior platform of a power plant, the device comprising:

[0011] An acquisition unit, used to acquire valid image frames of a target range, wherein the target range is an entrance management range of an inflammable and explosive area, and the valid image frames are used to represent continuous image frames in which a pedestrian target is selected by a boundary box;

[0012] A detection unit, configured to detect whether the pedestrian target triggers an unsafe behavior corresponding to the target range according to the valid image frame obtained by the acquisition unit, wherein the unsafe behavior is set according to the safety operation requirements of the flammable and explosive area;

[0013] An alarm unit is used to perform a corresponding alarm operation based on the behavior type of the pedestrian target triggering the unsafe behavior if the detection unit detects that the pedestrian target triggers the unsafe behavior corresponding to the target range according to the valid image frame.

[0014] In order to achieve the above-mentioned purpose, according to the third aspect of the present application, a storage medium is provided, which includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the unsafe behavior detection method for flammable and explosive areas based on the power plant unsafe behavior platform of the above-mentioned first aspect.

[0015] In order to achieve the above-mentioned purpose, according to the fourth aspect of the present application, a processor is provided, which is used to run a program, wherein the program, when running, executes the unsafe behavior detection method for flammable and explosive areas based on the power plant unsafe behavior platform of the above-mentioned first aspect.

[0016] By means of the above technical scheme, the present application provides a method, device, storage medium and processor for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant. When unsafe behavior detection is required for flammable and explosive areas, first obtain the valid image frame of the target range, the target range is the entrance management range of the flammable and explosive area, the valid image frame is used to characterize the continuous image frame with pedestrian targets selected by the boundary box, and then detect whether the pedestrian target triggers the unsafe behavior corresponding to the target range according to the valid image frame, the unsafe behavior is set according to the safe operation requirements of the flammable and explosive area, and finally determine if so, then perform the corresponding alarm operation based on the behavior type of the pedestrian target triggering the unsafe behavior. Through the technical scheme provided by the present application, it is possible to track the complex dynamic behavior of pedestrian targets based on continuous image frames with pedestrian targets, and detect whether the pedestrian targets trigger the unsafe behavior in the flammable and explosive area through the dynamic behavior, so as to accurately and comprehensively judge whether the staff meets the safe operation requirements of the flammable and explosive area, improve the detection accuracy of unsafe behavior in the flammable and explosive area, and effectively prevent accidents.

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0019] Figure 1 A flow chart of a method for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant provided in an embodiment of the present application is shown;

[0020] Figure 2 A flow chart of another unsafe behavior detection method for flammable and explosive areas based on an unsafe behavior platform of a power plant provided in an embodiment of the present application is shown;

[0021] Figure 3 A block diagram showing a composition of an unsafe behavior detection device for flammable and explosive areas based on an unsafe behavior platform of a power plant provided in an embodiment of the present application is shown;

[0022] Figure 4 A block diagram of another unsafe behavior detection device for flammable and explosive areas based on an unsafe behavior platform of a power plant provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0024] At present, the existing unsafe behavior detection in flammable and explosive areas based on the unsafe behavior platform of power plants usually relies on the behavior recognition of the monitoring system, that is, real-time capture of the corresponding video image frames, identification and judgment of whether the behavior in the image frame is unsafe behavior. However, this method is mostly based on the analysis of a single image frame, lacks consideration of the continuity of the action, and is difficult to accurately identify complex dynamic behaviors. The safety behavior specifications corresponding to flammable and explosive areas mostly need to last for a certain period of time. For example, the release of static electricity usually requires continuous contact with the static ball for more than 3 seconds, etc. This leads to low detection accuracy of unsafe behaviors in flammable and explosive areas, increasing the probability of accidents.

[0025] After research, the inventor found that since the safety behavior standards in flammable and explosive areas usually need to last for a certain period of time, it is possible to extract continuous video frames from the video stream at the entrance of the flammable and explosive area of ​​the pedestrian target to track the dynamic behavior of the pedestrian, and then compare the dynamic behavior with the unsafe behavior set according to the safety operation requirements of the flammable and explosive area, and execute the corresponding alarm. In this way, it is possible to accurately and comprehensively judge whether the staff meets the safety operation requirements of the flammable and explosive area, improve the detection accuracy of unsafe behavior in the flammable and explosive area, and effectively prevent accidents.

[0026] Based on the above considerations, the embodiment of the present application provides a method for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant. This method can improve the accuracy of detecting unsafe behavior in flammable and explosive areas and effectively prevent accidents. The specific execution steps are as follows: Figure 1 As shown, including:

[0027] 101. Obtain a valid image frame of a target range.

[0028] In this embodiment, the target range is the entrance management range of the inflammable and explosive area, and the effective image frame is used to characterize the continuous image frames in which the pedestrian target is selected by the bounding box. In practical applications, a high-definition surveillance camera that can monitor the entire management range will be installed at the entrance of the inflammable and explosive area, and the high-definition surveillance camera can be used to continuously and uninterruptedly collect video stream data of the target range. Continuous image frames containing pedestrian targets are extracted from the video stream through image analysis algorithms such as background subtraction, and then the pedestrian positions are accurately detected in the continuous image frames containing pedestrian targets through existing pedestrian detection models such as Yo l ov8, and marked in the form of a bounding box, so as to obtain continuous image frames in which the pedestrian target is selected by the bounding box. In addition, during the acquisition process, the video stream can also be preprocessed, including but not limited to removing noise, correcting light changes, etc., to ensure the quality of each image frame in the video stream.

[0029] 102. Detect whether a pedestrian target triggers an unsafe behavior corresponding to a target range according to a valid image frame.

[0030] In actual operations, workers have clear safety behavior standards before entering flammable and explosive areas, including but not limited to wearing safety helmets and work clothes correctly, completely discharging static electricity, and storing personal items that may cause fire or explosion (such as mobile phones and lighters). Therefore, in addition to installing high-definition surveillance cameras to monitor the entrance management range, related equipment such as static discharge balls and item storage boxes that are needed will be placed in different locations within the entrance management range based on safety behavior standards, as static ball areas and storage areas, so that the high-definition surveillance cameras can clearly and completely monitor the behavior of workers when they enter the static ball area or storage area.

[0031] In this embodiment, the unsafe behavior is set according to the safety operation requirements of the inflammable and explosive area. Since the safety behavior specifications for entering the inflammable and explosive area are clear, the corresponding safety operation requirements can be determined based on the safety behavior specifications. For example, the safety behavior specifications require thorough static discharge, and after combining the scale of inflammable and explosive gases in the inflammable and explosive area, in order to meet this specification and ensure the safety of the staff's work, it is necessary to ensure that the staff's hands are in continuous contact with the static discharge ball for at least 3 seconds, that is, the hands are in continuous contact with the static ball area for at least 3 seconds, and "being in the static ball area", "continuous contact" and "at least 3 seconds" are safety operation requirements.

[0032] Since unsafe behaviors may involve multiple aspects such as equipment wearing, isolation of prohibited items, and electrostatic discharge. Therefore, in order to ensure the accuracy of unsafe behavior detection and fully apply to flammable and explosive areas, corresponding detection models can be pre-trained based on multiple aspects such as equipment wearing, isolation of prohibited items or electrostatic discharge, such as tooling detection model, human key point detection model and item detection model, and corresponding evaluation criteria can be set, such as the simultaneous presence of long-sleeved work clothes and long-trouser work clothes is considered to be qualified for equipment wearing, the duration of the hand in the electrostatic ball area exceeds the preset duration is considered to be qualified for electrostatic discharge, and the presence of prohibited items and the movement trajectory of the hand in the storage area is a storage action. Among them, the key points include facial feature points, limb joints, etc., including but not limited to the nose, left and right eyes, left and right ears, left and right shoulders, left and right elbows, left and right wrists, left and right hips, left and right knees, left and right ankles, etc. For the capture of the trajectory or position of the above-mentioned continuous behavior of the key points of the hand or other key points, the target tracking algorithm such as ByteTrack can be used for continuous tracking. Through these models, the corresponding behaviors of equipment wearing, prohibited object isolation or electrostatic discharge are identified or tracked, and compared with the corresponding evaluation criteria to determine whether they trigger the unsafe behaviors corresponding to the target range. If not, no processing is performed. If triggered, step 103 is executed.

[0033] 103. Perform a corresponding warning operation based on the behavior type of the pedestrian target triggering the unsafe behavior.

[0034] Since it has been determined in step 102 that unsafe behavior has been triggered, in this step, classified alarms can be issued to pedestrian targets that have triggered unsafe behavior. For example, if it is found that a pedestrian target's work clothes are unqualified, the static discharge ball is not properly touched, or the prohibited items are not stored, the corresponding wear alarm operation, static alarm operation, or prohibited item alarm operation will be triggered. These alarm operations include but are not limited to real-time display on the large screen of the monitoring center or accompanied by sound broadcasts, etc., to ensure that pedestrian targets or relevant managers can quickly know and take measures to correct unsafe behaviors and effectively prevent accidents. In addition, the details of the alarm operation can also be recorded to provide data support for later safety training and accident analysis.

[0035] Based on the above Figure 1It can be seen from the implementation method that the present application provides a method for detecting unsafe behaviors in flammable and explosive areas based on the unsafe behavior platform of a power plant. When unsafe behaviors need to be detected in flammable and explosive areas, first obtain the valid image frame of the target range. The target range is the entrance management range of the flammable and explosive area. The valid image frame is used to characterize the continuous image frames in which the pedestrian target is selected by the boundary box. Then, according to the valid image frame, detect whether the pedestrian target triggers the unsafe behavior corresponding to the target range. The unsafe behavior is set according to the safety operation requirements of the flammable and explosive area. Finally, if it is determined that the pedestrian target triggers the unsafe behavior, the corresponding alarm operation is performed based on the behavior type of the pedestrian target triggering the unsafe behavior. Through the technical solution provided by the present application, it is possible to track the complex dynamic behaviors of pedestrian targets based on continuous image frames with pedestrian targets, and detect whether the pedestrian target triggers the unsafe behavior in the flammable and explosive area through the dynamic behavior, so as to accurately and comprehensively judge whether the staff meets the safety operation requirements of the flammable and explosive area, improve the detection accuracy of unsafe behaviors in the flammable and explosive area, and effectively prevent accidents.

[0036] Further, the preferred embodiment of the present application is in the above Figure 1 Based on this, a detailed description of the process of unsafe behavior detection in flammable and explosive areas based on the power plant unsafe behavior platform is given. The specific steps are as follows: Figure 2 As shown, including:

[0037] 201. Obtain a valid image frame of a target range.

[0038] This step is combined with the description of step 101 in the above method, and the same contents will not be repeated here.

[0039] 202. Capture a sub-image frame corresponding to the pedestrian target in the effective image frame according to the bounding box, and input the sub-image frame into a preset tooling detection model, a key point detection model, and an object detection model, respectively, to obtain a first detection result, a second detection result, and a third detection result.

[0040] Among them, the first detection result is the workwear category of the pedestrian target marked by the workwear detection model, the second detection result is the coordinates of multiple key points corresponding to the pedestrian target tracked by the key point detection model, and the third detection result is the category of the item carried by the behavioral target identified by the item detection model. In this step, the workwear detection model is specifically a multi-classification model, which can identify different types of workwear such as "short-sleeved non-workwear", "short-sleeved workwear", "long-sleeved non-workwear", "long-sleeved workwear", "shorts workwear", "shorts non-workwear", "long trousers workwear", "long trousers non-workwear" through training. The model is used to identify the type of workwear worn by the pedestrian target in the sub-image frame, and the identified work type is marked to obtain the first detection result. The key point detection model is specifically a Yo l ov8 target detection model trained based on the COCO-pose data set, which can identify 17 key point coordinates on the pedestrian target's body, such as facial feature points, limb joints, etc., specifically including nose, left and right eyes, left and right ears, left and right shoulders, left and right elbows, left and right wrists, left and right hips, left and right knees, and left and right ankles. By dynamically tracking the coordinates of these key points corresponding to the pedestrian target in the sub-image frame, the second detection result is obtained, and the second detection result can be used for subsequent dynamic action recognition, such as judging whether unsafe behaviors such as incomplete electrostatic discharge and failure to isolate prohibited items have been performed. The object detection model can be specifically trained by constructing an object data set based on actual scenarios such as thermal power plants. The items in the data set include but are not limited to mobile phones, lighters and other personal items that are prone to generate current or flames. The model is used to accurately identify the types of items carried around the hands of pedestrian targets, thereby obtaining a third detection result. The third detection result can be combined with the action recognition in the second detection result to comprehensively judge whether unsafe behaviors such as failure to isolate prohibited items have been performed.

[0041] 203. Determine whether the pedestrian target triggers the unsafe behavior corresponding to the equipment wearing based on the workwear category.

[0042] Based on the first detection result in step 202, that is, the workwear category of the pedestrian target can be determined, by comparing the workwear category with the preset workwear category standard, it can be determined whether the pedestrian target is wearing the workwear qualified, that is, whether the workwear that meets the safety operation requirements is correctly worn. The specific execution process is: based on the workwear category, it is determined whether the pedestrian target triggers the unsafe behavior corresponding to the equipment wearing, including: determining whether the workwear category meets the preset workwear category standard; if not, determining that the pedestrian target triggers the unsafe behavior corresponding to the equipment wearing.

[0043] Among them, the workwear category standard can be based on the specific safety requirements and industry standards of flammable and explosive areas, such as NFPA70E (American Electrical Safety Work Practice Standard), OSHA (Occupational Safety and Health Administration) regulations, etc., or it can be based on the potential hazards of static sparks, heat sources, chemical exposure, etc. in flammable and explosive areas. Risk assessment is performed to determine which type of workwear listed in step 202 can provide the best protection, thereby ensuring that the staff can minimize the risk of accidents. For example, in order to cover more skin area and reduce exposure, long-sleeved workwear and long-trousers workwear can be selected as the workwear category standard, or if it is determined that the risk is low after risk assessment and in order to ensure breathability, short-sleeved workwear and shorts workwear can be selected as the workwear category standard. This embodiment does not limit this.

[0044] 204. Determine whether a pedestrian target triggers an unsafe behavior corresponding to electrostatic discharge based on the coordinates of multiple key points.

[0045] Based on the second detection result in step 202, the coordinates of multiple key points corresponding to the dynamic behavior of the pedestrian target can be determined. By comparing whether the coordinates of multiple key points enter the electrostatic ball area within the target range, it is determined whether the electrostatic release ball is contacted, and combined with the duration of entering the electrostatic ball area within the target range, it is determined whether the pedestrian target has completely released static electricity. The specific execution process is: determine whether the duration of the target key point coordinates in the electrostatic ball area among the multiple key point coordinates exceeds the preset contact duration standard; if not, it is determined that the pedestrian target triggers an unsafe behavior corresponding to electrostatic release.

[0046] Among them, the electrostatic ball area is at the first position within the target range. Since the electrostatic ball area is pre-deployed within the target range, the first position has corresponding position coordinates, which can be all coordinates of the outer surface of the electrostatic release ball, or part of the coordinates of the contact position specified by the electrostatic release ball, which is not limited in this embodiment. Since electrostatic release is usually performed by touching the electrostatic release ball with the hand, the key points corresponding to the hand can be used as target key points, such as wrists, elbows, etc. By detecting whether the coordinates of the target key points coincide with the coordinates corresponding to the first position in space, it can be determined whether the pedestrian target contacts the electrostatic release ball. At the same time, since the effective image frames of this embodiment are continuous image frames, and there is an interval between adjacent image frames, the duration of the pedestrian target entering the electrostatic ball area can be counted according to the interval, that is, the duration of the pedestrian target continuously touching the electrostatic release ball. The duration is compared with the preset contact duration standard. If the duration exceeds the contact duration standard, it means that the pedestrian target has completely released static electricity. The contact duration standard is used to characterize the minimum duration of contact between the hand and the electrostatic release ball. Specifically, a reference duration can be set according to safety requirements, and then determined through measured data, such as the time required for static electricity release under different conditions such as dry environment or humid environment, to verify the reference duration, or a safety margin value can be superimposed on the reference time based on human experience, which is not limited in this embodiment. It is only necessary to ensure that static electricity is fully released to ensure the safety of flammable and explosive areas.

[0047] 205. Determine whether a pedestrian target triggers an unsafe behavior corresponding to electrostatic discharge based on multiple key point coordinates and object categories.

[0048] Based on the second detection result and the third detection result in step 202, it is possible to determine the coordinates of multiple key points corresponding to the dynamic behavior of the pedestrian target and the category of the items carried by the pedestrian target. By judging whether the item category is a prohibited item category such as a mobile phone, a lighter, etc., and then comparing the motion trajectory of the multiple key point coordinates entering the storage area position within the target range to determine whether it is a storage action of placing prohibited items in the storage box, it can be determined whether the pedestrian target has isolated the prohibited items. The specific execution process is: based on multiple key point coordinates and item categories, judging whether the pedestrian target triggers the unsafe behavior corresponding to the isolation of prohibited items, including: judging whether the item category does not belong to the preset prohibited item category, and whether the motion trajectory corresponding to the target key point coordinates in the multiple key point coordinates in the storage area meets the storage action standard; if not, it is determined that the pedestrian target triggers the unsafe behavior corresponding to the isolation of prohibited items.

[0049] Among them, the prohibited items category refers to the category of items that may cause fire, explosion or safety accidents. For example, electronic devices such as mobile phones, smart watches, tablet computers, etc. that are easy to generate electric current, or fire sources such as lighters, flints, magnesium rods, etc. that are easy to generate flames, etc., can be set specifically through industry standards or based on actual scene inspections. By comparing the determined item category with the prohibited item category, it can be determined whether it is a prohibited item. After determining that the item category is a prohibited item category, since the storage area is at the second position within the target range, and the storage area is also pre-deployed within the target range, the second position has corresponding position coordinates, which can be all coordinates of the outer surface of the entire storage box, or the specific coordinates of each grid of the storage box. This embodiment does not limit this. In addition, usually, items are carried or taken by hand, so the key points corresponding to the hand are also used as target key points, such as wrists, elbows, etc. By detecting whether the coordinates of the target key points coincide with the coordinates corresponding to the second position in space, it can be determined whether the pedestrian target touches the storage box, that is, whether it enters the storage area. When the pedestrian target enters the storage area, the hand movement trajectory can be determined by the change in the coordinates of the key points of its hand. By comparing the movement trajectory with the preset storage action standard, it can be determined that the pedestrian target has isolated the prohibited items. The storage action standard can be a specific action feature, such as stable placement, picking up, etc., or a specific position feature, such as the hand coordinates continue to stay in the storage area. It can also be compared whether there is a difference between the hand coordinates before and after entering and leaving the storage area, such as the presence of prohibited items before entering and the disappearance of prohibited items after exiting, etc. This embodiment does not limit this.

[0050] 206. Perform a corresponding alarm operation based on the behavior type of the unsafe behavior.

[0051] This step is combined with the description of step 103 in the above method, and the same content will not be repeated here. It should be noted that the specific execution process of executing the corresponding alarm operation based on the behavior type of unsafe behavior is: when the behavior type is unsafe behavior corresponding to workwear wearing, the wear alarm operation corresponding to the pedestrian target is triggered; when the behavior type is unsafe behavior corresponding to electrostatic discharge, the electrostatic alarm operation corresponding to the pedestrian target is triggered; when the behavior type is unsafe behavior corresponding to prohibited object isolation, the prohibited object alarm operation corresponding to the pedestrian target is triggered.

[0052] Among them, the wearable alarm operation, static electricity alarm operation and prohibited object alarm operation can be triggered according to specific pedestrian targets. For example, when a pedestrian target triggers an alarm operation, since the aforementioned steps also include the recognition of facial feature key points, the personal information of the pedestrian target can be extracted from the preset staff library through the facial feature key points, and the personal information and the corresponding triggered alarm operation can be combined to generate the corresponding alarm information, and the alarm information can be displayed in real time on the large screen of the monitoring center or accompanied by sound broadcast, etc. In addition, the alarm operation can also be executed through SMS, email or dedicated APP push. Through the classified alarms of the above three alarm methods, the accuracy of the alarm can be effectively improved, so that the pedestrian target or the relevant management personnel who trigger the corresponding unsafe behavior can be quickly informed and take measures to correct the unsafe behavior, effectively preventing the occurrence of accidents.

[0053] Furthermore, in order to enhance the safety behavior awareness in flammable and explosive areas, thereby reducing the occurrence of unsafe behaviors triggered by staff, it can also include: if the same pedestrian target continuously triggers at least two of the wear warning operation, static electricity warning operation and prohibited object warning operation, the pedestrian target will be marked as a dangerous target; according to the number of occurrences of the dangerous target, the behavior specification broadcast operation corresponding to the target range will be triggered.

[0054] Among them, since the electrostatic ball area and the storage area are respectively at the first position and the second position within the target range, that is to say, after entering the target range, the pedestrian target must pass through the first position and the second position in sequence to avoid the corresponding non-safe behaviors of non-release of static electricity or non-isolation of prohibited items, and the non-safe behavior detection of wearable work has been carried out after entering the target range. Therefore, the same pedestrian target may trigger multiple unsafe behaviors. Triggering multiple unsafe behaviors means that its safety behavior awareness is weak. Therefore, such pedestrian targets can be marked as dangerous targets after they appear. By counting the number of occurrences of dangerous targets, it is determined whether to trigger the behavior specification broadcast operation corresponding to the target range. The behavior specification broadcast operation is to broadcast the safety behavior specifications of flammable and explosive areas, thereby strengthening the safety awareness of all staff. Specifically, a periodic statistical or total statistical number threshold can be set in advance, and the behavior specification broadcast operation is determined by comparing with the preset number threshold.

[0055] Based on the above-mentioned description of the occurrence number triggering the behavior specification broadcast operation, specifically, if the occurrence number in the first cycle reaches the first preset number, the behavior specification broadcast operation is triggered and the occurrence number is cleared. Or if the occurrence number reaches the second preset number, the behavior specification broadcast operation is triggered according to the second cycle.

[0056] Among them, the first preset number of times can set a lower threshold as the trigger point for the initial warning, such as 3 times. When the number of appearances of the dangerous target within the first cycle (such as within a week) reaches the first preset number, a behavioral norms broadcast to all staff is automatically triggered. After the broadcast, the number of appearances of the dangerous target is automatically cleared and a new monitoring cycle begins. The second preset number of times can be set with a higher threshold, such as 20 times, indicating the total number of times unsafe behaviors are triggered. When the number of appearances of the dangerous target reaches the second preset number of times, the behavioral norms broadcast will be automatically played in a loop according to a predetermined period (such as every day or every week). Through the behavioral norms broadcast, the safety behavior awareness of staff entering flammable and explosive areas can be effectively strengthened, thereby further ensuring the safety of flammable and explosive areas and preventing accidents.

[0057] Furthermore, as a response to the above Figure 1-2 The implementation of the method embodiment shown in the figure, the embodiment of the present application provides a device for detecting unsafe behavior in flammable and explosive areas based on the unsafe behavior platform of a power plant, and the device is used to improve the accuracy of unsafe behavior detection in flammable and explosive areas and effectively prevent accidents. The embodiment of the device corresponds to the aforementioned method embodiment. For ease of reading, this embodiment will no longer repeat the details of the aforementioned method embodiment, but it should be clear that the device in this embodiment can correspond to all the contents of the aforementioned method embodiment. Specifically, Figure 3 As shown, the device comprises:

[0058] An acquisition unit 31 is used to acquire valid image frames of a target range, where the target range is an entrance management range of an inflammable and explosive area, and the valid image frames are used to represent continuous image frames in which a pedestrian target is selected by a boundary box;

[0059] A detection unit 32, configured to detect whether the pedestrian target triggers an unsafe behavior corresponding to the target range according to the valid image frame obtained by the acquisition unit 31, wherein the unsafe behavior is set according to the safety operation requirements of the flammable and explosive area;

[0060] The alarm unit 33 is configured to perform a corresponding alarm operation based on the behavior type of the unsafe behavior if the detection unit 32 detects that the pedestrian target triggers the unsafe behavior corresponding to the target range according to the valid image frame.

[0061] Further, such as Figure 4 As shown, the detection unit 32 includes:

[0062] The processing module 321 is used to intercept the sub-image frame corresponding to the pedestrian target in the effective image frame according to the bounding box, and input the sub-image frame into the preset tool detection model, key point detection model and object detection model respectively to obtain a first detection result, a second detection result and a third detection result, wherein the first detection result is the tool category of the pedestrian target marked by the tool detection model, the second detection result is the coordinates of multiple key points corresponding to the pedestrian target tracked by the key point detection model, and the third detection result is the category of the object carried by the behavior target identified by the object detection model;

[0063] A first detection module 322, configured to determine whether the pedestrian target triggers an unsafe behavior corresponding to the wearing of the equipment based on the workwear category obtained by the processing module 321;

[0064] A second detection module 323, configured to determine whether the pedestrian target triggers an unsafe behavior corresponding to the isolation of prohibited objects based on the coordinates of the multiple key points obtained by the processing module 321;

[0065] The third detection module 324 is used to determine whether the pedestrian target triggers an unsafe behavior corresponding to electrostatic discharge based on the multiple key point coordinates and the object category obtained by the processing module 321.

[0066] Further, such as Figure 4 As shown, the first detection module 322 is specifically used for:

[0067] Determining whether the tooling category meets the preset tooling category standard;

[0068] If not, it is determined that the pedestrian target triggers an unsafe behavior corresponding to the wearing of the equipment.

[0069] Further, such as Figure 4 As shown, the second detection module 323 is specifically used for:

[0070] Determine whether the duration of the target key point coordinates among the plurality of key point coordinates in the electrostatic ball area exceeds a preset contact duration standard, and the electrostatic ball area is at the first position within the target range;

[0071] If not, it is determined that the pedestrian target triggers the unsafe behavior corresponding to the electrostatic discharge.

[0072] Further, such as Figure 4 As shown, the third detection module 324 is specifically used for:

[0073] Determine whether the item category does not belong to the preset prohibited item category, and whether the motion trajectory corresponding to the target key point coordinates in the storage area among the multiple key point coordinates is a storage action, and the storage area is at a second position within the target range;

[0074] If not, it is determined that the pedestrian target triggers the unsafe behavior corresponding to the isolation of the prohibited objects.

[0075] Further, such as Figure 4 As shown, the alarm unit 33 includes:

[0076] The first alarm module 331 is used to trigger a wearing alarm operation corresponding to the pedestrian target when the behavior type is an unsafe behavior corresponding to the workwear wearing;

[0077] A second alarm module 332, configured to trigger an electrostatic alarm operation corresponding to the pedestrian target when the behavior type is an unsafe behavior corresponding to the electrostatic discharge;

[0078] The third alarm module 333 is used to trigger a prohibited object alarm operation corresponding to the pedestrian target when the behavior type is an unsafe behavior corresponding to the prohibited object isolation.

[0079] Further, such as Figure 4 As shown, the device also includes:

[0080] The marking module 334 is used to mark the pedestrian target as a dangerous target if the same pedestrian target continuously triggers at least two of the wearing warning operation, the static electricity warning operation and the prohibited object warning operation;

[0081] The broadcast module 335 is used to trigger the broadcast operation of the behavior specification corresponding to the target range according to the number of occurrences of the dangerous target obtained by the marking module 334.

[0082] Further, such as Figure 4 As shown, the broadcast module 335 is specifically used for:

[0083] If the number of occurrences in the first cycle reaches a first preset number, the behavior specification broadcast operation is triggered and the number of occurrences is cleared.

[0084] Further, such as Figure 4 As shown, the broadcast module 335 is specifically used for:

[0085] If the cumulative number of occurrences reaches a second preset number, the behavior specification broadcast operation is triggered in a second periodic cycle.

[0086] Furthermore, an embodiment of the present application further provides a storage medium, wherein the storage medium is used to store a computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the above Figure 1-2 The unsafe behavior detection method in flammable and explosive areas based on the unsafe behavior platform of the power plant is described in.

[0087] Furthermore, the embodiment of the present application also provides a processor, the processor is used to run a program, wherein the program executes the above Figure 1-2 The unsafe behavior detection method in flammable and explosive areas based on the unsafe behavior platform of the power plant is described in.

[0088] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0089] It is understandable that the related features in the above methods and devices can be referenced to each other. In addition, the "first", "second" and the like in the above embodiments are used to distinguish the embodiments, but do not represent the advantages and disadvantages of the embodiments.

[0090] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0091] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the application is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the application described herein, and the description of the specific language above is for the purpose of disclosing the best mode of implementation of the application.

[0092] In addition, the memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0093] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0094] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0095] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0096] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0097] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0098] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0099] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include transitory media such as modulated data signals and carrier waves.

[0100] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0101] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0102] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for detecting unsafe behavior in flammable and explosive areas based on an unsafe behavior platform of a power plant, characterized in that: The method comprises: Acquire valid image frames of a target range, where the target range is an entrance management range of an inflammable and explosive area, and the valid image frames are used to represent continuous image frames in which a pedestrian target is selected by a bounding box; Detecting, according to the valid image frame, whether the pedestrian target triggers an unsafe behavior corresponding to the target range, wherein the unsafe behavior is set according to the safety operation requirements of the flammable and explosive area; If so, a corresponding warning operation is performed based on the behavior type of the pedestrian target triggering the unsafe behavior.

2. The method according to claim 1, characterized in that Detecting whether the pedestrian target triggers an unsafe behavior corresponding to the target range according to the valid image frame includes: A sub-image frame corresponding to the pedestrian target is intercepted in the valid image frame according to the bounding box, and the sub-image frame is respectively input into a preset tooling detection model, a key point detection model and an object detection model to obtain a first detection result, a second detection result and a third detection result, wherein the first detection result is the tooling category of the pedestrian target marked by the tooling detection model, the second detection result is the coordinates of multiple key points corresponding to the pedestrian target tracked by the key point detection model, and the third detection result is the category of the object carried by the behavior target identified by the object detection model; Determining whether the pedestrian target triggers an unsafe behavior corresponding to the wearing of the equipment based on the type of the equipment; Determining whether the pedestrian target triggers an unsafe behavior corresponding to electrostatic discharge based on the coordinates of the plurality of key points; Based on the coordinates of the plurality of key points and the item category, it is determined whether the pedestrian target triggers an unsafe behavior corresponding to the isolation of prohibited items.

3. The method according to claim 2, characterized in that Judging whether the pedestrian target triggers an unsafe behavior corresponding to the wearing of the equipment based on the type of the equipment includes: Determining whether the tooling category meets the preset tooling category standard; If not, it is determined that the pedestrian target triggers an unsafe behavior corresponding to the wearing of the equipment.

4. The method according to claim 2, characterized in that: Judging whether the pedestrian target triggers an unsafe behavior corresponding to electrostatic discharge based on the coordinates of the plurality of key points includes: Determine whether the duration of the target key point coordinates among the plurality of key point coordinates in the electrostatic ball area exceeds a preset contact duration standard, and the electrostatic ball area is at the first position within the target range; If not, it is determined that the pedestrian target triggers an unsafe behavior corresponding to the electrostatic discharge.

5. The method according to claim 2, characterized in that: Determining whether the pedestrian target triggers an unsafe behavior corresponding to the isolation of prohibited objects based on the coordinates of the plurality of key points and the category of the object includes: Determine whether the item category does not belong to the preset prohibited item category, and whether the motion trajectory corresponding to the target key point coordinates in the storage area among the multiple key point coordinates meets the storage action standard, and the storage area is at the second position within the target range; If not, it is determined that the pedestrian target triggers the unsafe behavior corresponding to the isolation of the prohibited objects.

6. The method according to any one of claims 3 to 5, characterized in that: Executing a corresponding warning operation based on the behavior type of the pedestrian target triggering the unsafe behavior includes: When the behavior type is an unsafe behavior corresponding to the workwear wearing, triggering a wearing warning operation corresponding to the pedestrian target; When the behavior type is an unsafe behavior corresponding to the electrostatic discharge, triggering an electrostatic warning operation corresponding to the pedestrian target; When the behavior type is an unsafe behavior corresponding to the prohibited object isolation, a prohibited object warning operation corresponding to the pedestrian target is triggered.

7. The method according to claim 6, characterized in that The method further comprises: If the same pedestrian target continuously triggers at least two of the wearing warning operation, the static electricity warning operation and the prohibited object warning operation, the pedestrian target is marked as a dangerous target; The behavior specification broadcast operation corresponding to the target range is triggered according to the number of occurrences of the dangerous target.

8. A non-safe behavior detection device for flammable and explosive areas based on an unsafe behavior platform of a power plant, characterized in that: The device comprises: An acquisition unit, used to acquire valid image frames of a target range, wherein the target range is an entrance management range of an inflammable and explosive area, and the valid image frames are used to represent continuous image frames in which a pedestrian target is selected by a boundary box; A detection unit, configured to detect whether the pedestrian target triggers an unsafe behavior corresponding to the target range according to the valid image frame obtained by the acquisition unit, wherein the unsafe behavior is set according to the safety operation requirements of the flammable and explosive area; An alarm unit is used to perform a corresponding alarm operation based on the behavior type of the pedestrian target triggering the unsafe behavior if the detection unit detects that the pedestrian target triggers the unsafe behavior corresponding to the target range according to the valid image frame.

9. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the unsafe behavior detection method for flammable and explosive areas based on the power plant unsafe behavior platform as described in any one of claims 1 to 7.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the unsafe behavior detection method for flammable and explosive areas based on the unsafe behavior platform of a power plant as described in any one of claims 1 to 7.

Images