Reservoir dynamic monitoring system and method based on multi-source data

Abstract

The present invention relates to the field of intelligent monitoring technology, and specifically to a dynamic monitoring system and method for a reservoir based on multi-source data. The dynamic monitoring system for a reservoir based on multi-source data includes a monitoring acquisition module, which is used to obtain monitoring images taken by each front-end monitoring device; a monitoring identification module, which is used to identify each monitoring image, identify whether there is a person in the monitoring image, and when there is a person, identify the position where the person appears in the monitoring image, mark it as the monitoring position, and mark the monitoring image where the person appears as a detection image; an inspection identification module, which is preset with the inspection route of the staff, is used to analyze the position where the staff appears in the detection image according to the inspection route, and mark it as the inspection position; a picture analysis module, which is used to identify the deviation between the inspection position and the monitoring position, and when the deviation between the inspection position and the monitoring position meets the preset deviation condition, the person in the monitoring image is marked as a suspicious person.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to a reservoir dynamic monitoring system and method based on multi-source data. Background Art

[0002] Reservoirs are one of the widely used engineering measures for flood control in my country. Building reservoirs at appropriate locations in the upstream rivers of flood control areas can regulate and store flood water. Reservoirs are used to intercept and store flood water, reduce the peak flow entering the downstream rivers, and achieve the purpose of reducing flood disasters.

[0003] At present, especially in summer, ordinary people often enter reservoirs for fun. Although there are some reservoirs that allow ordinary people to enter for camping, fishing, swimming and other activities after the owner's permission. However, in order to prevent water pollution and protect the ecological environment, or during flood prevention, due to frequent strong weather such as rain and strong wind, the water level of mountain ponds and reservoirs has risen sharply, and there are great safety hazards in the water-related areas. In order to protect the safety of life and property, unrelated personnel are also prohibited from entering. Therefore, reservoirs that prohibit ordinary entry usually set up corresponding protective measures and set up monitoring to detect whether there are unrelated people breaking in. However, due to the large area occupied by the reservoir, it is difficult to achieve all-round monitoring of the reservoir area. There will always be certain blind spots, and it is impossible to ensure that the monitoring range covers every corner. At the same time, most of the illegal intruders break in at night. Reservoir monitoring is mostly carried out in night scenes. The light is dim at night. In the existing monitoring system, the method of judging whether the person in the monitoring screen is a staff member through face recognition is not applicable. Summary of the invention

[0004] The technical problem solved by the present invention is to provide a dynamic monitoring system for a reservoir based on multi-source data, which can improve the monitoring accuracy of whether the person is an intruder.

[0005] The basic solution provided by the present invention is: a dynamic monitoring system for a reservoir based on multi-source data, including a front-end monitoring device, a reservoir detection device and a server, wherein the server includes a monitoring acquisition module, a monitoring identification module, an inspection identification module, a location analysis module and a notification module;

[0006] Monitoring acquisition module, used to obtain monitoring images taken by various front-end monitoring devices;

[0007] The monitoring recognition module is used to identify each monitoring screen and identify whether there is a person in the monitoring screen. When a person appears, the position where the person appears in the monitoring screen is identified and marked as the monitoring position, and the monitoring screen where the person appears is marked as the detection screen;

[0008] The inspection recognition module has a preset inspection route for the staff, which is used to analyze the position where the staff appears in the detection screen according to the inspection route and mark it as the inspection position;

[0009] The image analysis module is used to identify the deviation distance between the inspection position and the monitoring position. When the deviation distance between the inspection distance and the monitoring position exceeds a preset distance threshold, the person in the monitoring image is marked as a suspicious person;

[0010] The notification module is used to identify the actual location of the suspicious person based on the monitored location and send the actual location to the staff.

[0011] The principle and advantage of the present invention are: in the reservoir area, each area is monitored according to the surveillance camera. When a person appears in the surveillance picture, the position where the surveillance person appears in the surveillance picture is identified. Here, identifying the surveillance picture means identifying the position of the person in the image in the canvas when the person first appears in the surveillance picture. Such as in the upper left corner, upper right corner, above, below, etc. of the surveillance picture. Thereafter, according to the preset inspection route of the inspection personnel, the position where the inspection personnel appear in the surveillance picture is identified. When the inspection personnel conduct inspections according to the inspection route, the positions where they appear in each surveillance picture are different, and the planned inspection routes are different at different times, and the positions where the inspection personnel appear in the surveillance picture are identified. If the distance difference between the positions is large, it means that the person is a suspicious person.

[0012] Compared with the prior art, most of the prior art is to identify the face and identify the identity information of the face to determine whether the person appearing in the monitoring screen is a staff member. This method has a low recognition accuracy in the case of night shooting, and when the face does not appear in the picture, it cannot be recognized. In this scheme, by identifying the position where the person appears in the monitoring screen, it is determined whether it is a patrol officer. When the reservoir staff conducts inspections, they usually conduct inspections according to a fixed route. When they conduct inspections according to a fixed inspection route, the position in the monitoring screen is relatively fixed. For example, when the inspection route comes from the left side of the monitoring location, the position where the patrol officer first appears in the monitoring screen is on the left side of the monitoring screen. At the same time, since the illegal intruders will try their best to avoid the patrol officers, the selected route will not overlap with the inspection route. When the illegal intruders appear in the monitoring screen, the position in the monitoring screen will have a large deviation from the position where the monitoring screen is entered according to the inspection route. In this way, the position where the person appears in the monitoring screen is identified to determine whether it is a suspicious person. Compared with the existing technology, there is no need to detect the face of the person, and it can also determine whether the person is a suspicious person even in dim lighting conditions.

[0013] Furthermore, it also includes a reservoir monitoring device, and the server also includes a reservoir collection module and a monitoring identification module;

[0014] Reservoir collection module, used to obtain various reservoir monitoring data collected by reservoir monitoring equipment;

[0015] The monitoring and identification module has preset risk thresholds for various reservoir monitoring data and identifies the difference between reservoir monitoring data and the risk thresholds;

[0016] The image analysis module is used to adjust the distance threshold according to the difference between the monitoring data and the analysis threshold. The smaller the difference between the monitoring data and the risk threshold, the smaller the distance threshold.

[0017] Through the reservoir monitoring equipment, various monitoring data in the reservoir, such as water level, water flow rate, etc., are identified. The smaller the difference between the monitoring data and the risk threshold, the smaller the distance threshold is when performing image analysis. That is, when the detection data is closer to the risk value, the identification adjustment of whether it is a suspicious person is more stringent. In this way, first, during the high-risk period of the reservoir, such as during the flood prevention period, it is strictly identified whether there are any illegal persons breaking in. Second, during the high-risk period of the reservoir, the requirements for patrol personnel are increased, and patrol personnel are required to strictly follow the planned patrol route for inspection. When deviation is found, they will receive a notification to avoid risks.

[0018] Furthermore, the picture analysis module includes a pixel marking module, a patrol analysis module, a first recognition module, and a second recognition module;

[0019] The pixel marking module is used to identify the image of the person at the monitoring location when a person appears in the picture, and to identify the position and number of pixels of the person's image;

[0020] Inspection analysis module, used to generate image detection frame and detection pixel number interval according to the inspection location and inspection route of staff;

[0021] The first recognition module is used to identify whether the number of pixels of the person image is within the detection pixel interval. If so, the deviation condition is not met, otherwise, the deviation condition is met;

[0022] The second recognition module is used to identify the number of pixels appearing in the picture detection frame according to the pixel positions of the person image when the deviation condition of the first recognition module is not met, and to identify whether the number of pixels appearing in the picture detection frame exceeds a preset number ratio. If so, the deviation condition is not met; if not, the deviation condition is met.

[0023] When a person image appears in the monitoring screen, the person image in the monitoring screen is first identified. The monitoring screen is in the form of a video image. The video image includes multiple pixels. For example, a 1080P video screen contains 3221054 pixels. When a person appears in the monitoring screen, the position and number of the pixels showing the person are identified. First, the number of pixels is identified. Since the staff enters the monitoring screen along a fixed route and the position is fixed, the size presented in the monitoring screen is fixed. The body shape of adults is similar. The number and position of the pixels displayed in the video image when the staff appears at a fixed position can be analyzed. If the number of pixels differs too much, it means that the deviation is large. Then, the number of person images appearing in the inspection frame is identified. If the number of images appearing in the inspection frame is small, it means that they do not appear in the monitoring screen along a fixed route. Since the video image is a two-dimensional screen, it is possible that the person appears at the same position in the monitoring screen at a different distance from the camera position. This problem can be avoided by detecting the number of pixels appearing in the inspection frame.

[0024] Furthermore, the picture analysis module also includes a third recognition module;

[0025] The inspection analysis module is also used to analyze the changes in the number of pixels in the monitoring screen of the staff according to the inspection route;

[0026] The third recognition module is used to identify the change in the number of pixels of the person image in the monitoring screen when the deviation condition of the second recognition module is not met, and determine whether it is consistent with the change in the number of pixels obtained by analysis. If they are consistent, the deviation condition is not met; if they are inconsistent, the deviation condition is met.

[0027] When the person in the monitoring screen moves, the image of the person in the monitoring screen will become larger or smaller in the two-dimensional image, and the number of pixels in the image of the person will change. Therefore, according to the inspection route of the patrol personnel when they appear in the monitoring screen, we can get where the patrol personnel enter the screen and where they leave the screen, so as to analyze the change in the number of pixels in the image of the person from entering the screen to leaving the screen. Then identify the change in the number of pixels of the person in the monitoring screen, and judge whether it is consistent with the change trend of the number of pixels appearing in the monitoring screen when the patrol personnel normally move along the inspection route. If it is inconsistent, it means that when the person appears in the monitoring screen, he does not move according to the inspection route, and the deviation condition is met. Through the recognition method of the change in the number of pixels, in the dim environment at night, as long as the outline of the person can be captured, it can be identified.

[0028] Furthermore, the server further comprises a multi-image detection module, and the multi-image detection module comprises a track recognition module and a route analysis module;

[0029] The trajectory recognition module is used to identify the movement trajectory of the suspicious person in the monitoring screen according to the position and quantity changes of the pixel points of the person's image in the monitoring screen when a person is judged as a suspicious person;

[0030] Route analysis module, used to analyze the action routes of suspicious persons based on their action trajectories in multiple surveillance images;

[0031] Notification module, used to send the analyzed action route to the staff.

[0032] The present invention also discloses a reservoir dynamic monitoring method based on multi-source data, which comprises the following steps:

[0033] S100: Acquire monitoring images captured by each front-end monitoring device;

[0034] S200: Identify each monitoring screen to identify whether there is a person in the monitoring screen. If there is a person, identify the position where the person appears in the monitoring screen, mark it as the monitoring position, and mark the monitoring screen where the person appears as the detection screen;

[0035] S300: according to the preset inspection route of the staff, the position where the staff appears in the detection picture is analyzed according to the inspection route, and marked as the inspection position;

[0036] S400: Identify the deviation between the inspection position and the monitoring position, and when the deviation between the inspection position and the monitoring position meets a preset deviation condition, mark the person in the monitoring image as a suspicious person;

[0037] S500: Identify the actual location of the suspicious person based on the monitored location, and send the actual location to the staff.

[0038] Further, the method further comprises the following steps:

[0039] S600: Acquire various reservoir monitoring data collected by reservoir monitoring equipment;

[0040] S700: According to the preset risk thresholds of various reservoir monitoring data, identifying the difference between the reservoir monitoring data and the risk threshold;

[0041] S800: According to the difference between the monitoring data and the analysis threshold, the deviation condition is adjusted. The smaller the difference between the monitoring data and the risk threshold, the lower the deviation condition.

[0042] Further, the S400 includes the following steps:

[0043] S401: When a person appears in the picture, an image of the person at the monitoring position is identified, and the position and number of pixel points of the image of the person are identified;

[0044] S402: Generate a picture detection frame and a detection pixel number interval according to the inspection position and the inspection route of the staff;

[0045] S403: Identify whether the number of pixels of the person's image is within the detection pixel interval. If so, the deviation condition is not met. Otherwise, the deviation condition is met.

[0046] S404: When the deviation condition of S403 is not met, the number of pixels appearing in the image detection frame is identified according to the pixel positions of the person image, and whether the number of pixels appearing in the image detection frame exceeds a preset number ratio. If so, the deviation condition is not met; if not, the deviation condition is met;

[0047] S405: Analyze the change in the number of pixels from the time when the staff member appears on the monitoring screen to the time when he leaves the monitoring screen according to the inspection route;

[0048] S406: When the deviation condition of S404 is not met, identify the change in the number of pixels of the person image in the monitoring screen, and determine whether it is consistent with the change in the number of pixels obtained by analysis. If they are consistent, the deviation condition is not met; if they are consistent, the deviation condition is met.

[0049] Further, the method further comprises the following steps:

[0050] S801: When a person is determined to be a suspicious person, the movement trajectory of the suspicious person in the monitoring screen is identified according to the position and quantity changes of the pixel points of the person's image in the monitoring screen;

[0051] S802: Analyze the action route of the suspicious person according to the action trajectory of the suspicious person in multiple monitoring images;

[0052] S803: Send the analyzed action route to the staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The present invention is a logic block diagram of an embodiment of a reservoir dynamic monitoring system based on multi-source data. DETAILED DESCRIPTION

[0054] The following is further described in detail through specific implementation methods:

[0055] The embodiment is basically as shown in the attached Figure 1 As shown:

[0056] A dynamic reservoir monitoring system based on multi-source data can improve the monitoring accuracy of whether there is an intruder.

[0057] The basic solution provided by the present invention is: a dynamic monitoring system for a reservoir based on multi-source data, including a front-end monitoring device, a reservoir detection device and a server, wherein the server includes a monitoring acquisition module, a monitoring identification module, an inspection identification module, a location analysis module and a notification module;

[0058] Monitoring acquisition module, used to obtain monitoring images taken by various front-end monitoring devices;

[0059] The monitoring recognition module is used to identify each monitoring screen and identify whether there is a person in the monitoring screen. When a person appears, the position where the person appears in the monitoring screen is identified and marked as the monitoring position, and the monitoring screen where the person appears is marked as the detection screen;

[0060] The inspection recognition module has a preset inspection route for the staff, which is used to analyze the position where the staff appears in the detection screen according to the inspection route and mark it as the inspection position;

[0061] The image analysis module is used to identify the deviation distance between the inspection position and the monitoring position. When the deviation distance between the inspection distance and the monitoring position exceeds a preset distance threshold, the person in the monitoring image is marked as a suspicious person;

[0062] The notification module is used to identify the actual location of the suspicious person based on the monitored location and send the actual location to the staff.

[0063] In the reservoir area, each area is monitored according to the surveillance cameras. When a person appears in the surveillance picture, the position where the surveillance person appears in the surveillance picture is identified. Here, identifying the surveillance picture means identifying the position of the person in the image in the canvas when the person first appears in the surveillance picture. Such as in the upper left corner, upper right corner, above, below, etc. of the surveillance picture. Thereafter, according to the preset inspection route of the inspection personnel, the position where the inspection personnel appear in the surveillance picture is identified. When the inspection personnel conduct inspections according to the inspection route, the positions where they appear in each surveillance picture are different, and the inspection routes planned at different times are different, and the positions where the inspection personnel appear in the surveillance picture are identified. If the distance difference between the positions is large, it means that the person is a suspicious person.

[0064] Compared with the prior art, most of the prior art is to identify the face and identify the identity information of the face to determine whether the person appearing in the monitoring screen is a staff member. This method has a low recognition accuracy in the case of night shooting, and when the face does not appear in the picture, it cannot be recognized. In this scheme, by identifying the position where the person appears in the monitoring screen, it is determined whether it is a patrol officer. When the reservoir staff conducts inspections, they usually conduct inspections according to a fixed route. When they conduct inspections according to a fixed inspection route, the position in the monitoring screen is relatively fixed. For example, when the inspection route comes from the left side of the monitoring location, the position where the patrol officer first appears in the monitoring screen is on the left side of the monitoring screen. At the same time, since the illegal intruders will try their best to avoid the patrol officers, the selected route will not overlap with the inspection route. When the illegal intruders appear in the monitoring screen, the position in the monitoring screen will have a large deviation from the position where the monitoring screen is entered according to the inspection route. In this way, the position where the person appears in the monitoring screen is identified to determine whether it is a suspicious person. Compared with the existing technology, there is no need to detect the face of the person, and it can also determine whether the person is a suspicious person even in dim lighting conditions.

[0065] It also includes a reservoir monitoring device, and the server also includes a reservoir collection module and a monitoring identification module;

[0066] Reservoir collection module, used to obtain various reservoir monitoring data collected by reservoir monitoring equipment;

[0067] The monitoring and identification module has preset risk thresholds for various reservoir monitoring data and identifies the difference between reservoir monitoring data and the risk thresholds;

[0068] The image analysis module is used to adjust the distance threshold according to the difference between the monitoring data and the analysis threshold. The smaller the difference between the monitoring data and the risk threshold, the smaller the distance threshold.

[0069] Through the reservoir monitoring equipment, various monitoring data in the reservoir, such as water level, water flow rate, etc., are identified. The smaller the difference between the monitoring data and the risk threshold, the smaller the distance threshold is when performing image analysis. That is, when the detection data is closer to the risk value, the identification adjustment of whether it is a suspicious person is more stringent. In this way, first, during the high-risk period of the reservoir, such as during the flood prevention period, it is strictly identified whether there are any illegal persons breaking in. Second, during the high-risk period of the reservoir, the requirements for patrol personnel are increased, and patrol personnel are required to strictly follow the planned patrol route for inspection. When deviation is found, they will receive a notification to avoid risks.

[0070] The picture analysis module includes a pixel marking module, an inspection analysis module, a first recognition module, and a second recognition module;

[0071] The pixel marking module is used to identify the image of the person at the monitoring location when a person appears in the picture, and to identify the position and number of pixels of the person's image;

[0072] Inspection analysis module, used to generate image detection frame and detection pixel number interval according to the inspection location and inspection route of staff;

[0073] The first recognition module is used to identify whether the number of pixels of the person image is within the detection pixel interval. If so, the deviation condition is not met, otherwise, the deviation condition is met;

[0074] The second recognition module is used to identify the number of pixels appearing in the picture detection frame according to the pixel positions of the person image when the deviation condition of the first recognition module is not met, and to identify whether the number of pixels appearing in the picture detection frame exceeds a preset number ratio. If so, the deviation condition is not met; if not, the deviation condition is met.

[0075] When a person image appears in the monitoring screen, the person image in the monitoring screen is first identified. The monitoring screen is in the form of a video image. The video image includes multiple pixels. For example, a 1080P video screen contains 3221054 pixels. When a person appears in the monitoring screen, the position and number of the pixels showing the person are identified. First, the number of pixels is identified. Since the staff enters the monitoring screen along a fixed route and the position is fixed, the size presented in the monitoring screen is fixed. The body shape of adults is similar. The number and position of the pixels displayed in the video image when the staff appears at a fixed position can be analyzed. If the number of pixels differs too much, it means that the deviation is large. Then, the number of person images appearing in the inspection frame is identified. If the number of images appearing in the inspection frame is small, it means that they do not appear in the monitoring screen along a fixed route. Since the video image is a two-dimensional screen, it is possible that the person appears at the same position in the monitoring screen at a different distance from the camera position. This problem can be avoided by detecting the number of pixels appearing in the inspection frame.

[0076] The picture analysis module also includes a third recognition module;

[0077] The inspection analysis module is also used to analyze the changes in the number of pixels in the monitoring screen of the staff according to the inspection route;

[0078] The third recognition module is used to identify the change in the number of pixel points of the person image in the monitoring screen when the deviation condition of the second recognition module is not met, and determine whether it is consistent with the change in the number of pixel points obtained by analysis. If they are consistent, the deviation condition is not met; if they are inconsistent, the deviation condition is met.

[0079] When the person in the monitoring screen moves, the image of the person in the monitoring screen will become larger or smaller in the two-dimensional image, and the number of pixels in the image of the person will change. Therefore, according to the inspection route of the patrol personnel when they appear in the monitoring screen, we can get where the patrol personnel enter the screen and where they leave the screen, so as to analyze the change in the number of pixels in the image of the person from entering the screen to leaving the screen. Then identify the change in the number of pixels of the person in the monitoring screen, and judge whether it is consistent with the change trend of the number of pixels appearing in the monitoring screen when the patrol personnel normally move along the inspection route. If it is inconsistent, it means that when the person appears in the monitoring screen, he does not move according to the inspection route, and the deviation condition is met. Through the recognition method of the change in the number of pixels, in the dim environment at night, as long as the outline of the person can be captured, it can be identified.

[0080] The server also includes a multi-image detection module, which includes a track recognition module and a route analysis module;

[0081] The trajectory recognition module is used to identify the movement trajectory of the suspicious person in the monitoring screen according to the position and quantity changes of the pixel points of the person's image in the monitoring screen when a person is judged as a suspicious person;

[0082] Route analysis module, used to analyze the action routes of suspicious persons based on their action trajectories in multiple surveillance images;

[0083] Notification module, used to send the analyzed action route to the staff.

[0084] The present invention also discloses a reservoir dynamic monitoring method based on multi-source data, which comprises the following steps:

[0085] S100: Acquire monitoring images captured by each front-end monitoring device;

[0086] S200: Identify each monitoring screen to identify whether there is a person in the monitoring screen. If there is a person, identify the position where the person appears in the monitoring screen, mark it as the monitoring position, and mark the monitoring screen where the person appears as the detection screen;

[0087] S300: according to the preset inspection route of the staff, the position where the staff appears in the detection picture is analyzed according to the inspection route, and marked as the inspection position;

[0088] S400: Identify the deviation between the inspection position and the monitoring position, and when the deviation between the inspection position and the monitoring position meets a preset deviation condition, mark the person in the monitoring image as a suspicious person;

[0089] S500: Identify the actual location of the suspicious person based on the monitored location, and send the actual location to the staff.

[0090] Further, the method further comprises the following steps:

[0091] S600: Acquire various reservoir monitoring data collected by reservoir monitoring equipment;

[0092] S700: According to the preset risk thresholds of various reservoir monitoring data, identifying the difference between the reservoir monitoring data and the risk threshold;

[0093] S800: According to the difference between the monitoring data and the analysis threshold, the deviation condition is adjusted. The smaller the difference between the monitoring data and the risk threshold, the lower the deviation condition.

[0094] The S400 includes the following steps:

[0095] S401: When a person appears in the picture, an image of the person at the monitoring position is identified, and the position and number of pixel points of the image of the person are identified;

[0096] S402: Generate a picture detection frame and a detection pixel number interval according to the inspection position and the inspection route of the staff;

[0097] S403: Identify whether the number of pixels of the person's image is within the detection pixel interval. If so, the deviation condition is not met. Otherwise, the deviation condition is met.

[0098] S404: When the deviation condition of S403 is not met, the number of pixels appearing in the image detection frame is identified according to the pixel positions of the person image, and whether the number of pixels appearing in the image detection frame exceeds a preset number ratio. If so, the deviation condition is not met; if not, the deviation condition is met;

[0099] S405: Analyze the change in the number of pixels from the time when the staff member appears on the monitoring screen to the time when he leaves the monitoring screen according to the inspection route;

[0100] S406: When the deviation condition of S404 is not met, identify the change in the number of pixels of the person image in the monitoring screen, and determine whether it is consistent with the change in the number of pixels obtained by analysis. If they are consistent, the deviation condition is not met; if they are consistent, the deviation condition is met.

[0101] The following steps are also included:

[0102] S801: When a person is determined to be a suspicious person, the movement trajectory of the suspicious person in the monitoring screen is identified according to the position and quantity changes of the pixel points of the person's image in the monitoring screen;

[0103] S802: Analyze the action route of the suspicious person according to the action trajectory of the suspicious person in multiple monitoring images;

[0104] S803: Send the analyzed action route to the staff.

[0105] The above are only embodiments of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme is not described in detail here. Ordinary technicians in the relevant field know all the common technical knowledge in the technical field to which the invention belongs before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply conventional experimental means before that date. Ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the scope of protection of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A reservoir dynamic monitoring system based on multi-source data, characterized by: It includes a front-end monitoring device and a server, wherein the server includes a monitoring acquisition module, a monitoring identification module, an inspection identification module, a location analysis module and a notification module; Monitoring acquisition module, used to obtain monitoring images taken by various front-end monitoring devices; The monitoring recognition module is used to identify each monitoring screen and identify whether there is a person in the monitoring screen. When a person appears, the position where the person appears in the monitoring screen is identified and marked as the monitoring position, and the monitoring screen where the person appears is marked as the detection screen; The inspection recognition module has a preset inspection route for the staff, which is used to analyze the position where the staff appears in the detection screen according to the inspection route and mark it as the inspection position; The image analysis module is used to identify the deviation between the inspection position and the monitoring position. When the deviation between the inspection position and the monitoring position meets the preset deviation condition, the person in the monitoring image is marked as a suspicious person; Notification module, used to identify the actual location of suspicious persons based on the monitored location and send the actual location to the staff; The picture analysis module includes a pixel marking module, an inspection analysis module, a first recognition module, and a second recognition module; The pixel marking module is used to identify the image of the person at the monitoring location when a person appears in the picture, and to identify the position and number of pixels of the person's image; Inspection analysis module, used to generate image detection frame and detection pixel number interval according to the inspection location and inspection route of staff; The first recognition module is used to identify whether the number of pixels of the person image is within the detection pixel interval. If so, the deviation condition is not met, otherwise, the deviation condition is met; The second recognition module is used to identify the number of pixels appearing in the picture detection frame according to the pixel positions of the person image when the deviation condition of the first recognition module is not met, and to identify whether the number of pixels appearing in the picture detection frame exceeds a preset number ratio. If so, the deviation condition is not met, and if not, the deviation condition is met; The server also includes a multi-image detection module, which includes a track recognition module and a route analysis module; The trajectory recognition module is used to identify the movement trajectory of the suspicious person in the monitoring screen according to the position and quantity changes of the pixel points of the person's image in the monitoring screen when a person is judged as a suspicious person; Route analysis module, used to analyze the action routes of suspicious persons based on their action trajectories in multiple surveillance images; Notification module, used to send the analyzed action route to the staff.

2. The reservoir dynamic monitoring system based on multi-source data according to claim 1 is characterized in that: It also includes a reservoir monitoring device, and the server also includes a reservoir collection module and a monitoring identification module; Reservoir collection module, used to obtain various reservoir monitoring data collected by reservoir monitoring equipment; The monitoring and identification module has preset risk thresholds for various reservoir monitoring data and identifies the difference between reservoir monitoring data and the risk thresholds; The image analysis module is used to adjust the deviation condition according to the difference between the monitoring data and the analysis threshold. The smaller the difference between the monitoring data and the risk threshold, the lower the deviation condition.

3. The reservoir dynamic monitoring system based on multi-source data according to claim 1 is characterized in that: The picture analysis module also includes a third recognition module; The inspection analysis module is also used to analyze the change in the number of pixels from the time when the staff member appears on the monitoring screen to the time when he leaves the monitoring screen according to the inspection route; The third recognition module is used to identify the change in the number of pixels of the person image in the monitoring screen when the deviation condition of the second recognition module is not met, and determine whether it is consistent with the change in the number of pixels obtained by analysis. If they are consistent, the deviation condition is not met; if they are inconsistent, the deviation condition is met.

4. A reservoir dynamic monitoring method based on multi-source data, characterized in that: The following steps are involved: S100: Acquire monitoring images captured by each front-end monitoring device; S200: Identify each monitoring screen to identify whether there is a person in the monitoring screen. If there is a person, identify the position where the person appears in the monitoring screen, mark it as the monitoring position, and mark the monitoring screen where the person appears as the detection screen; S300: according to the preset inspection route of the staff, the position where the staff appears in the detection picture is analyzed according to the inspection route, and marked as the inspection position; S400: Identify the deviation between the inspection position and the monitoring position, and when the deviation between the inspection position and the monitoring position meets a preset deviation condition, mark the person in the monitoring image as a suspicious person; S500: Identify the actual location of the suspicious person based on the monitored location, and send the actual location to the staff; The S400 includes the following steps: S401: When a person appears in the picture, an image of the person at the monitoring position is identified, and the position and number of pixel points of the image of the person are identified; S402: Generate a picture detection frame and a detection pixel number interval according to the inspection position and the inspection route of the staff; S403: Identify whether the number of pixels of the person's image is within the detection pixel interval. If so, the deviation condition is not met. Otherwise, the deviation condition is met. S404: When the deviation condition of S403 is not met, the number of pixels appearing in the image detection frame is identified according to the pixel positions of the person image, and whether the number of pixels appearing in the image detection frame exceeds a preset number ratio. If so, the deviation condition is not met; if not, the deviation condition is met; S405: Analyze the change in the number of pixels from the time when the staff member appears on the monitoring screen to the time when he leaves the monitoring screen according to the inspection route; S406: When the deviation condition of S404 is not met, identify the change in the number of pixels of the person image in the monitoring screen, and determine whether it is consistent with the change in the number of pixels obtained by analysis. If they are consistent, the deviation condition is not met, and if they are consistent, the deviation condition is met; The following steps are also included: S801: When a person is determined to be a suspicious person, the movement trajectory of the suspicious person in the monitoring screen is identified according to the position and quantity changes of the pixel points of the person's image in the monitoring screen; S802: Analyze the action route of the suspicious person according to the action trajectory of the suspicious person in multiple monitoring images; S803: Send the analyzed action route to the staff.

5. The reservoir dynamic monitoring method based on multi-source data according to claim 4 is characterized in that: The following steps are also included: S600: Acquire various reservoir monitoring data collected by reservoir monitoring equipment; S700: According to the preset risk thresholds of various reservoir monitoring data, identifying the difference between the reservoir monitoring data and the risk threshold; S800: According to the difference between the monitoring data and the analysis threshold, the deviation condition is adjusted. The smaller the difference between the monitoring data and the risk threshold, the lower the deviation condition.

Images