Monitoring and early warning method, device, equipment and storage medium

Abstract

The present application discloses a monitoring and early warning method, device, equipment and storage medium, which relates to the field of image recognition technology, and includes: acquiring monitoring images, obtaining person information and environmental information based on the monitoring and early warning monitoring images; conducting environmental hazard assessment based on the monitoring and early warning environmental information to obtain an environmental hazard level; extracting facial features and limb features based on the monitoring and early warning person information, conducting person abnormality analysis based on the facial features and limb features to obtain a person hazard level; obtaining an early warning type based on the monitoring and early warning person hazard level and the environmental hazard level, and issuing an early warning based on the monitoring and early warning type; the method improves the early warning and danger prevention capabilities of the security monitoring system by comprehensively analyzing the person information and environmental information in the monitoring images, and at the same time improves the accuracy of abnormal behavior recognition, solving the problem that danger cannot be recognized and warned in a timely manner when it appears in the current monitoring screen.

Description

Technical Field

[0001] The present application relates to the field of image recognition technology, and in particular to monitoring and early warning methods, devices, equipment and storage media. Background Art

[0002] Modern society has an urgent need for security management and crime prevention. With the acceleration of urbanization, frequent public security incidents and the rapid development of information technology, traditional security monitoring methods can no longer meet the growing security requirements.

[0003] Early video surveillance systems mainly relied on manual monitoring, which was not only inefficient, but also prone to omission of important information due to human negligence. With the surge in the number of surveillance cameras, it has become almost impossible to manually review all recordings. In recent years, the development of big data technology has made it possible to store and analyze massive amounts of surveillance data. Artificial intelligence, especially the application of deep learning technology, enables the system to automatically identify specific objects, behavior patterns, and abnormal situations in the video, significantly improving the level of intelligent monitoring. However, the breadth and depth of monitoring use are currently not being improved, and it is impossible to identify dangers and issue warnings in a timely and accurate manner when dangerous phenomena appear on the surveillance screen.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide a monitoring and early warning method, device, equipment and storage medium, aiming to solve the technical problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, they cannot be identified in time and early warning of the danger cannot be issued.

[0006] To achieve the above objectives, the present application proposes a monitoring and early warning method, which includes:

[0007] Acquire a surveillance image, and obtain person information and environment information based on the surveillance image;

[0008] Performing an environmental hazard assessment based on the environmental information to obtain an environmental hazard level;

[0009] Extracting facial features and limb features based on the character information, performing character abnormality analysis according to the facial features and the limb features, and obtaining the character danger level;

[0010] The warning type is obtained according to the human danger level and the environmental danger level, and monitoring and warning are performed based on the warning type.

[0011] In one embodiment, performing an environmental risk assessment based on the environmental information to obtain an environmental risk level includes:

[0012] Obtaining a current environment category according to the environmental information, wherein the current environment category is a normal environment, a potentially dangerous environment, and a risky environment;

[0013] When the environment category is a normal environment, the environmental hazard level is obtained according to the preset risk level;

[0014] When the environment category is a potentially hazardous environment, identifying potential hazardous factors in the environmental information, and obtaining an environmental hazard level according to the potential hazardous factors;

[0015] When the environment category is a risk environment, risk characteristics are obtained according to the risk environment, risk assessment is performed based on the risk characteristics to obtain assessment results, and the environmental hazard level is obtained according to the assessment results.

[0016] In one embodiment, when the environment category is a normal environment, obtaining the environmental hazard level according to a preset risk level includes:

[0017] When the environment category is a normal environment, generating a picture detection signal so that the monitoring image generates a preset picture according to the picture detection signal;

[0018] When the preset picture is detected, the preset risk level is used as the environmental hazard level;

[0019] When the preset picture is not detected within the preset time, the timestamp of the historical picture detection signal is obtained, the level correction coefficient is obtained according to the timestamp and the current time, and the environmental hazard level is obtained according to the level correction coefficient and the preset risk level.

[0020] In one embodiment, extracting facial features and limb features based on the character information, performing character abnormality analysis according to the facial features and the limb features, and obtaining the character danger level includes:

[0021] Extracting facial features and body features based on the character information;

[0022] Matching the facial features with the facial features of preset dangerous persons to obtain a matching degree;

[0023] Performing motion analysis on the limb features to obtain an initial motion risk level;

[0024] Evaluate the limb movement based on reference environment information to obtain a risk correction coefficient;

[0025] The character danger level is obtained according to the danger correction coefficient, the initial action danger and the matching degree.

[0026] In one embodiment, the performing motion analysis on the limb features to obtain an initial motion risk level includes:

[0027] Extract skeleton joints according to the limb features to obtain initial behavior actions;

[0028] Generate a skeleton sequence graph based on the initial behavior action, and perform spatial convolution on the skeleton sequence spatiotemporal graph to obtain spatial features;

[0029] Performing temporal convolution on the spatial features to obtain temporal features;

[0030] A predicted behavior is obtained based on the spatial feature and the temporal feature, and the predicted behavior is evaluated to obtain an initial action risk.

[0031] In one embodiment, the evaluating the limb movement based on the reference environment information to obtain the risk correction coefficient includes:

[0032] Based on the reference environmental information, the reference number of people and environmental harshness factors are obtained;

[0033] Obtaining a risk correction coefficient based on the reference number of people and the environmental adverse factors;

[0034] The obtaining of the character danger level according to the danger correction coefficient, the initial action danger and the matching degree includes:

[0035] The initial action risk is adjusted according to the risk correction coefficient to obtain a reference risk;

[0036] The reference danger level is adjusted according to the matching degree to obtain the danger level of the person.

[0037] In one embodiment, obtaining the warning type according to the person danger level and the environmental danger level, and performing monitoring and warning based on the warning type, includes:

[0038] When the person danger level is greater than or equal to the person danger level threshold and the environment danger level is greater than or equal to the environment danger level threshold, compare the person danger level and the environment danger level to obtain a comparison result, and obtain a reference danger level according to the comparison result;

[0039] When the person danger level is less than the person danger level threshold and the environment danger level is greater than or equal to the environment danger level threshold, the environment danger level is used as a reference danger level;

[0040] Matching the reference hazard level with a preset hazard level table to obtain a warning type, wherein the preset hazard level table includes one-to-one corresponding hazard levels and warning types;

[0041] Monitoring and warning are performed based on the warning type.

[0042] In addition, to achieve the above purpose, the present application also proposes a monitoring and early warning device, which includes:

[0043] A parameter acquisition module, used to acquire a monitoring image, and obtain person information and / or environmental information based on the monitoring image;

[0044] A level assessment module, used to perform an environmental risk assessment based on the environmental information to obtain an environmental risk level;

[0045] The level assessment module is further used to extract facial features and limb features based on the character information, perform character abnormality analysis based on the facial features and limb features, and obtain the character's danger level;

[0046] The monitoring and early warning module is used to perform monitoring and early warning according to the personal danger level and / or environmental danger level.

[0047] In addition, to achieve the above-mentioned purpose, the present application also proposes a monitoring and early warning device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the monitoring and early warning method described above.

[0048] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the monitoring and early warning method described above are implemented.

[0049] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the monitoring and early warning method described above.

[0050] One or more technical solutions proposed in the present application have at least the following technical effects: through comprehensive analysis of the personal information and environmental information in the monitoring image, the danger assessment capability of the monitoring screen is enhanced, the early warning and danger prevention capabilities of the security monitoring system are improved, and at the same time, the accuracy of abnormal behavior recognition is improved and the possibility of false alarms and missed alarms is reduced, which solves the problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, they cannot be identified in time and the danger cannot be warned. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0053] Figure 1 A flowchart of the first embodiment of the monitoring and early warning method of this application is provided;

[0054] Figure 2 A flow chart of the second embodiment of the monitoring and early warning method of this application is provided;

[0055] Figure 3 A schematic diagram of character information feature extraction provided in an embodiment of the monitoring and early warning method of this application;

[0056] Figure 4 This is a schematic diagram of the module structure of the monitoring and early warning device according to an embodiment of the present application;

[0057] Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the monitoring and early warning method in the embodiment of the present application. DETAILED DESCRIPTION

[0058] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0059] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0060] The main solution of the embodiment of the present application is: acquire monitoring images, obtain character information and environmental information based on the monitoring warning monitoring images; perform environmental hazard assessment based on the monitoring warning environmental information to obtain the environmental hazard level; extract facial features and limb features based on the monitoring warning character information, perform character abnormality analysis based on the facial features and limb features to obtain the character hazard level; obtain the warning type based on the monitoring warning character hazard level and the environmental hazard level, and perform monitoring warning based on the monitoring warning type.

[0061] In this embodiment, for ease of description, the following description is made with identification of monitoring and early warning equipment as the execution subject.

[0062] Due to the existing technology, modern society has an urgent need for security management and crime prevention. With the acceleration of urbanization, frequent public safety incidents and the rapid development of information technology, traditional security monitoring methods can no longer meet the growing security requirements. Early video surveillance systems mainly relied on manual monitoring, which was not only inefficient, but also prone to omission of important information due to human negligence. With the surge in the number of surveillance cameras, it has become almost impossible to manually view all recordings. In recent years, the development of big data technology has made it possible to store and analyze massive surveillance data. Artificial intelligence, especially the application of deep learning technology, enables the system to automatically identify specific objects, behavior patterns and abnormal situations in the video, significantly improving the level of intelligent monitoring. However, the breadth and depth of monitoring use cannot be improved, and it is impossible to identify dangers and issue warnings in a timely and accurate manner when dangerous phenomena appear on the monitoring screen.

[0063] The present application provides a solution that improves the early warning and danger prevention capabilities of the security monitoring system, while also improving the accuracy of abnormal behavior recognition, solving the current problem of being unable to timely identify and issue early warnings when danger appears in the monitoring screen.

[0064] It can be seen from the above embodiments that the present application discloses a monitoring and early warning method, device, equipment and storage medium, which relates to the field of image recognition technology, and discloses the following: acquiring monitoring images, and obtaining character information and environmental information based on the monitoring and early warning monitoring images; performing environmental hazard assessment based on the monitoring and early warning environmental information to obtain an environmental hazard level; extracting facial features and limb features based on the monitoring and early warning character information, and performing character abnormality analysis based on the facial features and limb features to obtain a character hazard level; obtaining an early warning type based on the monitoring and early warning character hazard level and the environmental hazard level, and performing monitoring and early warning based on the monitoring and early warning type; the method improves the early warning and danger prevention capabilities of the security monitoring system by comprehensively analyzing the character information and environmental information in the monitoring images, and at the same time improves the accuracy of abnormal behavior recognition, solving the problem that danger cannot be recognized and warned in a timely manner when it occurs in the current monitoring screen.

[0065] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a monitoring and early warning device, etc. The following takes the monitoring and early warning device as an example to illustrate this embodiment and the following embodiments.

[0066] Based on this, the present application embodiment provides a monitoring and early warning method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the monitoring and early warning method of the present application.

[0067] In this embodiment, the monitoring and early warning method includes steps S10 to S40:

[0068] Step S10, acquiring a surveillance image, and obtaining person information and environment information based on the surveillance image.

[0069] It is understandable that the surveillance image may be any frame of image captured from a surveillance video, or may be a surveillance image acquired periodically, or may be a surveillance image obtained by other means.

[0070] It should be noted that character information may include facial features, action features, posture features, and expression features; environmental information may include object information, scene information, and weather information.

[0071] It should be emphasized that item information may include controlled knives, hammers, axes and other dangerous items; scene information may include fire, flood, leakage, collapse, damage, failure, normal and other scenes; weather information may include rain, snow, thunder, strong wind and other weather categories.

[0072] Step S20: performing an environmental hazard assessment based on the environmental information to obtain an environmental hazard level.

[0073] It is understandable that the environmental hazard levels may include no hazard, level one hazard, level two hazard, level three hazard, and level four hazard.

[0074] It should be noted that the risk assessment based on environmental information can be a comprehensive assessment of the environmental risk level by integrating the object information, scene information and weather information in the environmental information.

[0075] In a feasible implementation, step S20 may include steps A21 to A24:

[0076] Step A21, obtaining the current environment category according to the environmental information, wherein the current environment category is a normal environment, a potentially dangerous environment, and a risky environment.

[0077] It is understandable that the current environment type can be obtained according to the environmental information by extracting features from the monitoring image to obtain background features, and then performing feature recognition based on the background features to obtain the vehicles, objects, plants, rain / sunny / snow / fog / flood / fire / thunder, road surface, wall, animals, etc. in the background.

[0078] It should be understood that the environmental information can be obtained by extracting features from the surveillance image.

[0079] In the specific implementation, the surveillance image is subjected to preprocessing operations such as denoising, grayscale, and normalization before feature extraction to obtain environmental information features to improve the accuracy of feature extraction; then the training data is collected and labeled, including positive samples (foreground-person information-person features) and negative samples (background-environmental information-environmental features); then a deep learning architecture is built based on U-Net, Mask R-CNN or a model based on the attention mechanism, and then the model is trained using the labeled data set, and the trained model is applied to new images or video frames to directly output the segmentation results of the background and foreground or the object bounding box, and the environmental features are obtained based on the object bounding box.

[0080] Step A22, when the environment category is a normal environment, obtain the environmental hazard level according to the preset risk level.

[0081] It is understandable that the preset risk level may be no risk.

[0082] It should be understood that when the environment category is a normal environment, it can be understood that the current environment does not have environmental characteristics such as vehicle collisions, vehicle fires, plant collapse, heavy rain, heavy snow, heavy fog, water accumulation, fire, wall falling off, road damage, road collapse, animal gathering, animal frenzy, etc.

[0083] It should be noted that when the environment category is a normal environment, the environmental hazard level is obtained according to the preset risk level, including: when the environment category is a normal environment, a picture detection signal is generated so that the monitoring image produces a preset picture according to the picture detection signal; when the preset picture is detected, the preset risk level is used as the environmental hazard level; when the preset picture is not detected within a preset time, the timestamp of the historical picture detection signal is obtained, the level correction coefficient is obtained according to the timestamp and the current time, and the environmental hazard level is obtained according to the level correction coefficient and the preset risk level.

[0084] Among them, it can be understood that the picture detection signal can be a signal automatically generated when it is detected that the current environment category is a normal environment. Based on this signal, the controller can send a preset picture to the monitoring device. The preset picture can be a network disconnection picture caused by the signal stopping, or it can be a specific picture, which can be adjusted according to actual conditions.

[0085] Among them, it should be understood that when the environment category is a normal environment, generating a picture detection signal can make a preset picture appear in the monitoring image. Based on the preset picture, it can be effectively detected whether the monitoring device is in normal operation. If the preset picture appears, it is considered to be operating normally. If the preset picture does not appear, it is considered that the monitoring device may have a fault, and there may be problems with the currently identified normal environment.

[0086] Among them, it should be noted that the historical picture detection signal can be understood as the picture detection signal corresponding to the successful acquisition of the preset picture in the monitoring device in the past; in simple terms, the timestamp of obtaining the historical picture detection signal can be the time point when the picture detection signal was generated when the preset picture was successfully detected last time.

[0087] Among them, it should be emphasized that the timestamp of the last successful detection of the preset screen is compared with the timestamp of the current time to see the length of the time interval, and the preset danger level is adjusted based on the length of the time interval.

[0088] Among them, it can be understood that the longer the interval time is, the longer the monitoring equipment has failed, and the higher the possible danger is. The time interval is obtained according to the timestamp and the current time, the level correction coefficient is obtained according to the time interval, and the environmental hazard level is obtained according to the level correction coefficient and the preset risk level.

[0089] In the specific implementation, if the time interval is less than one day, the level correction coefficient is 0; if the time interval is greater than or equal to 1 day and less than 3 days, the correction coefficient is 1; if the time interval is greater than or equal to 3 days and less than or equal to 7 days, the correction coefficient is 2; the correction coefficient is added to the preset risk level to obtain the environmental hazard level.

[0090] Step A23, when the environment category is a potentially dangerous environment, identifying potential risk factors in the environmental information, and obtaining an environmental risk level according to the potential risk factors.

[0091] It is understandable that a potentially dangerous environment can be understood as the existence of relatively severe weather characteristics in the current environment, such as heavy rain, heavy snow, heavy fog, and accumulated water; or the existence of severe animal-related characteristics, such as animal gathering, animal frenzy, etc.; or the existence of relatively severe weather characteristics and severe animal-related characteristics at the same time.

[0092] It should be understood that the environmental hazard level may include no hazard, level one hazard, level two hazard, level three hazard and level four hazard. When there are relatively severe weather characteristics or there are severe animal-related characteristics, the environmental hazard level is level one hazard; if there are both relatively severe weather characteristics and animal-related severe characteristics, the environmental hazard level is level two hazard.

[0093] Step A24, when the environment category is a risk environment, obtain risk characteristics according to the risk environment, perform risk assessment based on the risk characteristics, obtain assessment results, and obtain the environmental hazard level according to the assessment results.

[0094] It is understandable that when the current environment has special adverse features such as vehicle collisions, vehicle fires, plant collapse, wall falling, road collapse, etc., the current environment category is considered to be a risk environment.

[0095] It should be emphasized that the environmental features corresponding to different environmental categories can be added to the environmental features corresponding to each environmental category according to the usage during subsequent use. For example, it can be added that when falling objects from high altitude are identified, the current environmental category is considered to be a risky environment.

[0096] It should be understood that the risk assessment is performed based on the risk characteristics, and the assessment result can be to conduct a risk assessment on the risk environment according to the urgency of the environmental characteristics corresponding to the risk environment, for example, vehicle collision and vehicle fire are particularly urgent; plant falling, wall falling, and road collapse are generally urgent.

[0097] It should be emphasized that the environmental level corresponding to the particularly urgent risk environment is set at Level 4, and the environmental level corresponding to the general urgent risk environment is set at Level 3.

[0098] It is understandable that the classification of different environmental characteristics can be achieved based on actual conditions or other judgment methods, and this embodiment does not limit this.

[0099] In this implementation, by adopting different assessment strategies in different environments, risk assessment is made more accurate, which not only ensures the efficiency of daily management, but also improves the pertinence and accuracy of the assessment by identifying specific risk factors or risk characteristics in potential danger or risk environments, ensuring that the classification of environmental hazard levels is more scientific and reasonable.

[0100] The above are only feasible implementations of step S20 provided in this embodiment, and this embodiment does not specifically limit the specific implementation of step S20.

[0101] Step S30, extracting facial features and limb features based on the character information, performing character abnormality analysis based on the facial features and the limb features, and obtaining the character's danger level.

[0102] It is understandable that the extraction of facial features can be used to identify dangerous persons, such as wanted criminals, habitual thieves, etc. on the one hand; on the other hand, it can be used to judge the emotional state of the person under surveillance in combination with facial features.

[0103] It should be understood that the limb features can be extracted to determine whether an accident has occurred to a person in the surveillance video based on the limb features. Dangerous situations such as falling to the ground, electric shock, collision, and a child being trapped can be analyzed through limb features.

[0104] It should be noted that the personal danger levels may include no personal danger, first-level personal danger, second-level personal danger, third-level personal danger and fourth-level personal danger.

[0105] It should be emphasized that the analysis of a person's danger level can be divided into two types. One is to identify dangerous persons based on facial features, and obtain the first person's danger level based on the identification results; then obtain facial emotions based on facial features, obtain the person's behavior status based on limb features, and obtain the second person's danger level based on facial emotions and the person's behavior status; and then obtain the person's danger level based on the first person's danger level and the second person's danger level.

[0106] It is understandable that the first person danger level can be divided into first-level dangerous person and normal; the second person danger level can be divided into normal, first-level person danger and second-level person danger.

[0107] It should be understood that when judging the danger level of a person, facial features may or may not be recognized. If facial features do not exist, the danger level is judged directly based on the person's body movement features.

[0108] It should be understood that the first person danger level and the second person danger level may exist at the same time, or only one of the two may exist, and the existing danger level will be used as the person danger level.

[0109] Step S40, obtaining a warning type according to the person danger level and the environmental danger level, and performing monitoring and warning based on the warning type.

[0110] It is understandable that the warning types can be divided into voice reminders, rescue reminders, and alarm reminders, and different warning types correspond to different monitoring warnings.

[0111] It should be understood that monitoring warnings of different warning types may also be adjusted according to actual conditions, and this embodiment does not limit this.

[0112] In a feasible implementation, step S40 may include steps A41 to A44:

[0113] Step A41, when the human danger level is greater than or equal to the human danger level threshold and the environmental danger level is greater than or equal to the environmental danger level threshold, compare the human danger level and the environmental danger level to obtain a comparison result, and obtain a reference danger level based on the comparison result.

[0114] It is understandable that the character level threshold can be understood as the lowest character level that is set to affect the danger level. Similarly, the environment level threshold can be understood as the lowest environment level that is set to affect the danger level.

[0115] It should be noted that the comparison result is obtained by comparing the human danger level and the environmental danger level. The reference danger level obtained according to the comparison result may be to use the higher of the human danger level and the environmental danger level as the reference danger level.

[0116] Step A42: When the human danger level is less than the human danger level threshold and the environmental danger level is greater than or equal to the environmental danger level threshold, the environmental danger level is used as a reference danger level.

[0117] It is understandable that when the danger level of a person is less than the person level threshold, it can be understood that the danger level of the person is not used to consider the danger level at this time.

[0118] It should be emphasized that when the person danger level is greater than or equal to the person danger level threshold and the environment danger level is less than the environment danger level threshold, the person danger level is used as the reference danger level.

[0119] It should be further emphasized that when the character danger level is less than the character level threshold and the environment danger level is less than the environment level threshold, the preset danger level is used as the reference danger level, where the preset danger level can be no danger, level 1 danger level, etc., which is a pre-set danger level. The preset danger level can also be adjusted according to actual conditions, and this embodiment does not limit this.

[0120] It should be noted that there may not be a human danger level in the monitoring environment, but there will be an environmental danger level, just like there may not always be people in the monitoring but there will always be a background environment.

[0121] Step A43, matching the reference danger level with a preset danger level table to obtain a warning type, wherein the preset danger level table includes one-to-one corresponding danger levels and warning types.

[0122] It is understandable that the preset hazard level table can refer to Table 1, as follows:

[0123]

[0124] It should be understood that the reference hazard level may be one of the hazard levels. Based on the hazard level corresponding to the reference hazard level, the corresponding warning type in the preset hazard level table may be obtained, and then monitoring and warning may be performed based on the warning type.

[0125] Step A44, performing monitoring and warning based on the warning type.

[0126] In specific implementation, the monitoring warning corresponding to the voice reminder can be a reminder that there is a dangerous hazard here, and maintenance is required in time, or a voice safety reminder can be given to the people around the monitoring; if the monitoring warning corresponding to the rescue reminder is a reminder that assistance is needed here, and staff must be dispatched as soon as possible to perform maintenance and deal with the danger, and at the same time, reminders and warnings can be given to the people around the monitoring through voice; if the monitoring warning corresponding to the alarm reminder is an alarm processing, a voice reminder of the principle monitoring area.

[0127] In this implementation, by simultaneously considering personal behavior risks (person danger level) and surrounding environment risks (environmental danger level), we not only focus on the degree of danger of the individual itself, but also take into account the impact of the external environment on its safety, so that the warning is more in line with the actual situation, and improve the effectiveness and pertinence of the warning. By setting person level thresholds and environmental level thresholds, we allow flexible adjustment according to actual conditions to adapt to different safety needs and risk tolerances, and can effectively balance the sensitivity and false alarm rate of the alarm, ensuring that warnings are issued in situations that really need attention to avoid unnecessary interference.

[0128] The above are only feasible implementations of step S40 provided in this embodiment, and this embodiment does not specifically limit the specific implementation of step S40.

[0129] This embodiment provides a monitoring and early warning method, which enhances the danger assessment capability of the monitoring screen through comprehensive analysis of the personal information and environmental information in the monitoring image, improves the early warning and danger prevention capabilities of the security monitoring system, and at the same time improves the accuracy of abnormal behavior recognition and reduces the possibility of false alarms and missed alarms, solving the problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, they cannot be identified in time and the danger cannot be warned.

[0130] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 2 , step S30, the monitoring and early warning method further includes steps S31~S35:

[0131] Step S31, extracting facial features and body features based on the character information.

[0132] It is understandable that the person information can be obtained by using object detection technology to locate the position of the person in the image, and selecting the area containing the person. The specific object detection technology can be selected according to actual conditions, and this embodiment does not limit this.

[0133] It should be understood that extracting facial features and limb features based on character information may be to perform body key point detection to obtain limb features and to perform retrograde facial detection to obtain facial information, wherein the limb features include feature point information of various joints of the human body.

[0134] It should be noted that the character information in this embodiment is obtained based on the Swin-Transformer gaze tracking algorithm, which can be referred to Figure 3, the Swin-Transformer gaze tracking algorithm in the figure adopts a dual-channel method. When input into the network structure for sharing scene features, the scene and head information can be completely independent. So in this structure, no matter how many people are in the scene, the scene image is independently input into the scene channel for one-time feature extraction. For video understanding based on surveillance cameras, the scene image basically rarely changes. The dual-channel structure can avoid the scene image from being repeatedly extracted and increasing the amount of calculation. As for the head channel, it is related to the number of people on the scene. Everyone who needs to predict the gaze target will be cropped to obtain a head image, and a head position binary mask will be generated according to the position of the head in the scene image. In each scene, different head features share scene features, and each head image and its position mask will be placed in the head channel separately to obtain more accurate facial features.

[0135] Step S32, matching the facial features with the preset facial features of dangerous persons to obtain a matching degree.

[0136] It is understandable that the facial features of preset dangerous persons can be accessed through judicial authorization. When law enforcement agencies handle cases, they can enter the facial information of confirmed dangerous persons into a specific database through judicial procedures and authorization. These databases are only used by legally authorized institutions within the scope prescribed by law for crime prevention, investigation and other purposes.

[0137] It should be noted that the facial features are matched with the facial features of the preset dangerous persons, and the matching degree can be obtained by extracting the key feature points of the face, such as the positions of the eyes, nose, and mouth, through algorithms such as Active Shape Model (ASM), Active Appearance Model (AAM), or more modern deep learning methods such as Face Alignment Networks (FANs). Then, the positions and relative distances of these feature points in the two facial images are compared to evaluate their matching degree.

[0138] It should be further explained that the facial feature matching degree can also be calculated by using deep learning models (such as VGGFace, FaceNet, ArcFace, etc.) to convert facial images into high-dimensional feature vectors. These vectors can capture the unique features of the face, and in the vector space, the feature vectors of similar faces are close to each other. The similarity between these feature vectors is calculated. Commonly used methods include Euclidean distance, cosine similarity or metric learning-based methods. The smaller the distance, the more similar the two faces are.

[0139] In a specific implementation, the steps of facial feature collection may include collecting facial feature samples; screening and extracting important facial features, and importing the extracted data into our feature database; performing identity authentication, and matching the collected facial feature information in the database according to a specific matching algorithm. When the matching degree reaches a certain value, it indicates that the recognition is successful.

[0140] Step S33, performing motion analysis on the limb features to obtain an initial motion risk level.

[0141] It is understandable that limb features include multiple joint features. By analyzing multiple joint features, character movements are obtained, and the character's behavior type is judged based on the movement, such as walking, squatting, leaning, falling, getting stuck, etc. Based on the behavior type, it is further analyzed whether the movement poses a safety risk to obtain the initial movement danger level.

[0142] In a feasible implementation, step S33 may include steps A331 to A334:

[0143] Step A331, extracting skeleton joints according to the limb features to obtain initial behavior actions.

[0144] It is understandable that skeleton joint extraction can be to use a human body detection algorithm to detect the position of the human body in each frame, usually outputting a human body frame and a confidence score; and then using an advanced human posture estimation model to locate the key joints of the human body.

[0145] It should be understood that the human body detection algorithm can be a professional human body detection model such as YOLOv4; the human body posture estimation model can be a model such as OpenPose, COCO-POSE Estimation, HRNet, etc.

[0146] It should be noted that

[0147] Step A332, generating a skeleton sequence graph based on the initial behavior action, and performing spatial convolution on the skeleton sequence spatiotemporal graph to obtain spatial features.

[0148] It is understandable that, according to the initial behavior action, the key joints may include the neck, shoulders, elbows, wrists, hips, knees, ankles, etc., and a skeleton sequence diagram of the human body is formed based on each key joint.

[0149] It should be noted that the spatial convolution can refer to the following formula:

[0150]

[0151] Among them, A represents the adjacency matrix of the virtual-real combined spatial convolution, E is the degree of the adjacency matrix A subset, M is the weight of the edge of the adjacency matrix A subset, W is the weight function of the joint point, and f1 represents the spatial feature.

[0152] Furthermore, spatial convolution of the skeleton sequence spatiotemporal graph can be expressed as f1(G), where G represents the skeleton sequence spatiotemporal graph.

[0153] Step A333, performing time convolution on the spatial feature to obtain the time feature.

[0154] It should be noted that the temporal convolution can refer to the following formula.

[0155]

[0156] Among them, F in represents the input feature, that is, the spatial feature, P represents the joint point of the limb feature, and W represents the weight function.

[0157] Furthermore, temporal convolution of spatial features can be expressed as f2(f1(G)).

[0158] Step A334, obtaining a predicted behavior based on the spatial features and the temporal features, and evaluating the predicted behavior to obtain an initial action risk.

[0159] It should be emphasized that the predicted behavior based on the spatial features and the temporal features can refer to the following formula:

[0160]

[0161] Among them, f out represents the body movement category, f1 represents the virtual-real combination spatial convolution, and f2 represents the temporal convolution.

[0162] Furthermore, the corresponding initial action danger level is obtained according to the physical action category. For example, if the physical action category is walking, the initial action danger level is 0; if the physical action category is squatting, leaning, or falling, the initial danger level is 1; if the physical action category is getting stuck, the danger level is 2; and if the physical action category is fighting, the danger level is 3.

[0163] In this implementation, by generating a spatiotemporal graph of the skeleton sequence and performing spatial convolution and temporal convolution respectively, the motion pattern is deeply analyzed from two dimensions: the spatial features capture the relative position and form of the limbs in the spatial layout, while the temporal convolution reveals the trend and speed of the movement evolution over time, which helps to accurately identify ongoing behaviors and predict upcoming actions, thereby improving the accuracy and timeliness of behavior recognition.

[0164] The above are only feasible implementations of step S33 provided in this embodiment, and this embodiment does not specifically limit the specific implementation of step S33.

[0165] Step S34, evaluating the limb movement based on the reference environment information to obtain a risk correction coefficient.

[0166] It is understandable that the reference environment information may include weather type and number of people.

[0167] It should be understood that weather types include sunny, cloudy, indoor, light rain, rainstorm, strong wind, light snow, etc. Weather types are divided into normal weather and bad weather. The specific division method can be adjusted according to actual conditions, and this embodiment does not limit this.

[0168] It should be noted that the environmental severity factor corresponding to normal weather is 1, and the environmental severity factor corresponding to severe weather is 2, which can also be adjusted according to actual conditions.

[0169] It should be noted that the reference number of people and the environmental severity factor are obtained based on the reference environmental information, and the risk correction coefficient is obtained based on the reference number of people and the environmental severity factor.

[0170] It should be further explained that in the current environment, too few people will have a certain impact on safety, and too many people will also have a certain impact on the corresponding number. A range of people is set. If the reference number of people falls within the range, the number impact factor is 1. If the reference number of people is less than the range, the number impact factor is set to 1.1. If the reference number of people is greater than the range, the number impact factor is set to 1.5.

[0171] It is understandable that the number of people influence factor is obtained based on the reference number of people, and the risk correction coefficient is obtained by multiplying the number of people influence factor and the environmental severity factor.

[0172] It should be noted that the setting of the number of people and the influencing factors can be adjusted according to actual conditions, and this embodiment does not limit this.

[0173] Step S35, obtaining the character danger level according to the danger correction coefficient, the initial action danger level and the matching degree.

[0174] It should be understood that the initial action risk is adjusted according to the risk correction coefficient to obtain a reference risk.

[0175] It should be noted that the matching degree ranges from 1 to 100. The matching degree is normalized and then multiplied by the reference danger level to adjust the reference matching degree and obtain the character danger level.

[0176] This embodiment provides a monitoring and early warning method. Through advanced face recognition technology, facial features are accurately extracted from character information and matched with a preset database of dangerous persons. Potential threat persons can be quickly identified, which improves the pertinence and efficiency of monitoring. Through detailed analysis of body features, initial movements that may indicate dangerous behavior can be identified, and reference environmental information is introduced to evaluate body movements, and the risk correction coefficient is calculated, which means that the technology can adjust the evaluation criteria according to specific situations, reducing the possibility of false positives and missed positives. Combined with facial feature matching, initial movement risk, and environmental impact risk correction coefficient, multi-dimensional evaluation can provide more comprehensive and dynamic security assessment results, making it easier for security managers to take corresponding countermeasures according to different risk levels, from warnings, enhanced monitoring to direct intervention, to effectively prevent and control the occurrence of potential security incidents.

[0177] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the monitoring and early warning method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0178] This application also provides a monitoring and early warning device, please refer to Figure 4 , the monitoring and early warning device comprises:

[0179] A parameter acquisition module 10 is used to acquire a monitoring image and obtain person information and / or environmental information based on the monitoring image;

[0180] A level assessment module 20, configured to perform an environmental risk assessment based on the environmental information to obtain an environmental risk level;

[0181] The level assessment module 20 is further used to extract facial features and limb features based on the character information, perform character abnormality analysis based on the facial features and limb features, and obtain the character's danger level;

[0182] The monitoring and warning module 30 is used to perform monitoring and warning according to the person danger level and / or environmental danger level.

[0183] The monitoring and early warning device provided by the present application adopts the monitoring and early warning method in the above embodiment, which can solve the technical problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, it cannot be identified in time, and the danger cannot be warned. Compared with the prior art, the beneficial effects of the monitoring and early warning device provided by the present application are the same as the beneficial effects of the monitoring and early warning method provided by the above embodiment, and the other technical features in the monitoring and early warning device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0184] The present application provides a monitoring and early warning device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the monitoring and early warning method in the above-mentioned embodiment one.

[0185] Reference below Figure 5 , which shows a schematic diagram of the structure of a monitoring and early warning device suitable for implementing the embodiment of the present application. The monitoring and early warning device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The monitoring and early warning device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0186] like Figure 5 As shown, the monitoring and early warning device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the monitoring and early warning device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the monitoring and early warning device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a monitoring and early warning device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0187] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0188] The monitoring and early warning device provided by the present application adopts the monitoring and early warning method in the above embodiment, which can solve the technical problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, it cannot be identified in time, and the danger cannot be warned. Compared with the prior art, the beneficial effects of the monitoring and early warning device provided by the present application are the same as the beneficial effects of the monitoring and early warning method provided by the above embodiment, and the other technical features in the monitoring and early warning device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0189] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0190] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0191] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the monitoring and early warning method in the above-mentioned embodiment.

[0192] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0193] The above-mentioned computer-readable storage medium may be included in the monitoring and early warning device; or it may exist independently without being assembled into the monitoring and early warning device.

[0194] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the monitoring and early warning device, the monitoring and early warning device: obtains monitoring images, and obtains character information and environmental information based on the monitoring and early warning monitoring images; performs environmental hazard assessment based on the monitoring and early warning environmental information to obtain the environmental hazard level; extracts facial features and limb features based on the monitoring and early warning character information, performs character abnormality analysis based on the monitoring and early warning facial features and the monitoring and early warning limb features to obtain the character hazard level; obtains the warning type based on the monitoring and early warning character hazard level and the environmental hazard level, and performs monitoring and early warning based on the monitoring and early warning type.

[0195] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0196] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0197] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0198] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned monitoring and early warning method, and can solve the technical problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, it cannot be identified in time, and the danger cannot be warned. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the monitoring and early warning method provided by the above-mentioned embodiment, and will not be repeated here.

[0199] The present application also provides a computer program product, including a computer program, which implements the steps of the monitoring and early warning method as described above when executed by a processor.

[0200] The computer program product provided by this application can solve the technical problem that the current monitoring content requires human supervision, and if dangerous phenomena appear in the monitoring content, it cannot be identified in time, and the danger cannot be warned. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the monitoring and warning method provided by the above embodiment, which will not be repeated here.

[0201] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A monitoring and early warning method, characterized in that: The monitoring and early warning method comprises: Acquire a surveillance image, and obtain person information and environment information based on the surveillance image; Performing an environmental hazard assessment based on the environmental information to obtain an environmental hazard level; Extracting facial features and limb features based on the character information, performing character abnormality analysis according to the facial features and the limb features, and obtaining the character danger level; Obtaining a warning type according to the danger level of the person and the danger level of the environment, and performing monitoring and warning based on the warning type; The step of performing an environmental risk assessment based on the environmental information to obtain an environmental risk level includes: Obtaining a current environment category according to the environmental information, wherein the current environment category is a normal environment, a potentially dangerous environment, and a risky environment; When the environment category is a normal environment, the environmental hazard level is obtained according to the preset risk level; When the environment category is a potentially hazardous environment, identifying potential hazardous factors in the environmental information, and obtaining an environmental hazard level according to the potential hazardous factors; When the environment category is a risk environment, a risk feature is obtained according to the risk environment, a risk assessment is performed based on the risk feature to obtain an assessment result, and an environmental hazard level is obtained according to the assessment result; When the environment category is a normal environment, the environmental hazard level is obtained according to the preset risk level, including: When the environment category is a normal environment, generating a picture detection signal so that the monitoring image generates a preset picture according to the picture detection signal; When the preset picture is detected, the preset risk level is used as the environmental hazard level; When the preset picture is not detected within the preset time, the timestamp of the historical picture detection signal is obtained, the level correction coefficient is obtained according to the timestamp and the current time, and the environmental hazard level is obtained according to the level correction coefficient and the preset risk level.

2. The monitoring and early warning method according to claim 1, characterized in that: The extracting facial features and limb features based on the character information, performing character abnormality analysis according to the facial features and the limb features, and obtaining the character danger level includes: Extracting facial features and body features based on the character information; Matching the facial features with the facial features of preset dangerous persons to obtain a matching degree; Performing motion analysis on the limb features to obtain an initial motion risk level; Evaluate the limb movement based on reference environment information to obtain a risk correction coefficient; The character danger level is obtained according to the danger correction coefficient, the initial action danger and the matching degree.

3. The monitoring and early warning method according to claim 2, characterized in that: The step of performing motion analysis on the limb features to obtain an initial motion risk level includes: Extract skeleton joints according to the limb features to obtain initial behavior actions; Generate a skeleton sequence graph based on the initial behavior action, and perform spatial convolution on the skeleton sequence spatiotemporal graph to obtain spatial features; Performing temporal convolution on the spatial features to obtain temporal features; A predicted behavior is obtained based on the spatial feature and the temporal feature, and the predicted behavior is evaluated to obtain an initial action risk.

4. The monitoring and early warning method according to claim 2, characterized in that: The step of evaluating the limb movement based on the reference environment information to obtain a risk correction coefficient includes: Based on the reference environmental information, the reference number of people and environmental harshness factors are obtained; Obtaining a risk correction coefficient based on the reference number of people and the environmental adverse factors; The obtaining of the character danger level according to the danger correction coefficient, the initial action danger and the matching degree includes: The initial action risk is adjusted according to the risk correction coefficient to obtain a reference risk; The reference danger level is adjusted according to the matching degree to obtain the danger level of the person.

5. The monitoring and early warning method according to claim 1, characterized in that: The step of obtaining a warning type according to the danger level of the person and the danger level of the environment, and performing monitoring and warning based on the warning type includes: When the person danger level is greater than or equal to the person danger level threshold and the environment danger level is greater than or equal to the environment danger level threshold, compare the person danger level and the environment danger level to obtain a comparison result, and obtain a reference danger level according to the comparison result; When the person danger level is less than the person danger level threshold and the environment danger level is greater than or equal to the environment danger level threshold, the environment danger level is used as a reference danger level; Matching the reference hazard level with a preset hazard level table to obtain a warning type, wherein the preset hazard level table includes one-to-one corresponding hazard levels and warning types; Monitoring and warning are performed based on the warning type.

6. A monitoring and early warning device, characterized in that: The monitoring and early warning device comprises: A parameter acquisition module, used to acquire a monitoring image, and obtain person information and environment information based on the monitoring image; A level assessment module, used to perform an environmental risk assessment based on the environmental information to obtain an environmental risk level; The level assessment module is further used to extract facial features and limb features based on the character information, perform character abnormality analysis based on the facial features and limb features, and obtain the character's danger level; A monitoring and warning module, used to obtain a warning type according to the danger level of the person and the danger level of the environment, and to perform monitoring and warning based on the warning type; The level assessment module is further used to obtain the current environment category according to the environmental information, the current environment category being a normal environment, a potential dangerous environment, and a risky environment; when the environment category is a normal environment, the environmental danger level is obtained according to a preset risk level; when the environment category is a potential dangerous environment, the potential risk factors in the environmental information are identified, and the environmental danger level is obtained according to the potential risk factors; when the environment category is a risky environment, risk characteristics are obtained according to the risky environment, risk assessment is performed based on the risk characteristics, an assessment result is obtained, and the environmental danger level is obtained according to the assessment result; The level assessment module is also used to generate a picture detection signal when the environment category is a normal environment, so that the monitoring image produces a preset picture according to the picture detection signal; when the preset picture is detected, the preset risk level is used as the environmental hazard level; when the preset picture is not detected within a preset time, the timestamp of the historical picture detection signal is obtained, the level correction coefficient is obtained according to the timestamp and the current time, and the environmental hazard level is obtained according to the level correction coefficient and the preset risk level.

7. A monitoring and early warning device, characterized in that: The device comprises: a memory, a processor, and a monitoring and early warning program stored in the memory and executable on the processor, wherein the monitoring and early warning program is configured to implement the monitoring and early warning method according to any one of claims 1 to 5.

8. A storage medium, characterized in that: The storage medium stores a monitoring and early warning program, and when the monitoring and early warning program is executed by the processor, the monitoring and early warning method according to any one of claims 1 to 5 is implemented.

Images