CN115604424A - Material storage security protection management system - Google Patents
Material storage security protection management system Download PDFInfo
- Publication number
- CN115604424A CN115604424A CN202211102853.7A CN202211102853A CN115604424A CN 115604424 A CN115604424 A CN 115604424A CN 202211102853 A CN202211102853 A CN 202211102853A CN 115604424 A CN115604424 A CN 115604424A
- Authority
- CN
- China
- Prior art keywords
- target
- frame
- personnel
- equipment
- target detection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/44—Event detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/24—Reminder alarms, e.g. anti-loss alarms
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Computer Interaction (AREA)
- Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Social Psychology (AREA)
- Signal Processing (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Warehouses Or Storage Devices (AREA)
Abstract
The application provides a material storage security protection management system. The storage system comprises a storage area, a storage module, a personnel monitoring module, a storage module and a device monitoring module, wherein the storage module, the personnel monitoring module and the device monitoring module are mutually independent, and the storage module is used for respectively identifying, detecting and tracking a material target, a personnel target and a device target in the storage area, so that a pixel moving track of a target detection frame is converted into a corresponding three-dimensional coordinate through a storage model, and moving states and operation actions of the materials, the personnel and the device are accurately extracted. Therefore, the safety protection module can trigger the alarm prompt to ensure the safety of the warehouse when the corresponding operation exceeds the authority or is not in accordance with the operation standard. Because this application can synthesize the running state who judges personnel, equipment, goods and materials, consequently, can also further pass through the model training and detecting and accord with authority requirement and operating specification, trigger warehouse the control unit linkage and let pass in order to improve storage management transportation efficiency during the operation of no safety risk.
Description
Technical Field
The application relates to the technical field of warehousing management, in particular to a material warehousing security management system.
Background
The existing warehouse system is complex and redundant, consists of a plurality of traditional monitoring devices, and is integrally represented as single function, independent system, complex data, low utilization rate, and lack of systematic resource integration and intelligent management analysis.
Moreover, the conventional material management is difficult. The reason is that the traditional material supervision method is difficult to be applied to an automatic three-dimensional warehouse, the storage state of materials cannot be checked in real time, trays can only be taken out of the warehouse by means of a stacker, or manual adventure enters a roadway to be checked, the efficiency is low, and safety risks exist.
In addition, the existing storage security technology has the defect that the intelligent level is backward. The existing warehouse operation management mainly adopts a people's air defense supervision mode, cannot realize the security protection management effect of all-round and all-time coverage, and derives the problems of human resource waste, supervision loopholes and the like, thereby increasing the warehouse operation risk.
Disclosure of Invention
This application provides a goods and materials storage security protection management system to prior art's not enough, and this application combines the storage model who types in advance through the motion state of synthesizing tracking personnel, goods and materials equipment, can accurately realize surmounting the operation action that the authority and be not conform to operating specification to warehouse goods and materials, discernment surmount the authority to and the record is tracked in time to the suggestion of reporting to the police and realization. The technical scheme is specifically adopted in the application.
Firstly, in order to achieve the above object, a material storage security management system is provided, which comprises: the cameras are arranged at a plurality of positions of the storage area and are used for shooting video images of the storage area; the material monitoring module is used for identifying, detecting and tracking a material target moved out of the image shelf area according to the video images of the storage area shot by each camera; the personnel monitoring module is used for identifying personnel targets in the storage area according to the storage area video images shot by the cameras, detecting and tracking the personnel targets and determining the identities of the personnel targets; the equipment monitoring module is used for identifying, detecting and tracking equipment targets at channel positions among the shelf areas according to the storage area video images shot by the cameras; the storage model is connected with the material monitoring module, the personnel monitoring module and the equipment monitoring module and used for calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to pixel positions of the material target, the personnel target and the equipment target under different camera shooting angles; and the security and protection verification module is connected with the storage model and used for determining the moving state and the operation action of the storage model according to the specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model, triggering a reaction terminal arranged near the target to give an alarm prompt and store the tracking record of each camera when the operation exceeds the authority or does not accord with the operation specification, and triggering a corresponding storage control unit arranged near the target to pass when the operation does not exceed the authority and accords with the operation specification.
Optionally, in the material warehousing security management system as described in any one of the above, each of the material monitoring module, the personnel monitoring module, and the equipment monitoring module is respectively and independently provided with the following operation units to identify, detect, and track the monitored target: the target detection unit is used for identifying detection targets contained in each frame of the video image by utilizing a pre-trained yolov3 model, correspondingly extracting target detection frames corresponding to materials, personnel or equipment, and marking identification numbers corresponding to the target detection frames; a Kalman filtering unit for generating a frame prediction frame according to each target detection frame in the previous frame; the Hungarian algorithm unit respectively identifies the association condition before the target detection frame in each frame according to the matching degree of the pixel coordinate positions between the target detection frame and the prediction frame of the frame, and marks the mutually associated target detection frames in each frame as having the same identification number; the matching track unit is used for respectively remapping target detection frames shot by different cameras into the visual angle ranges of other cameras according to the visual angle ranges shot by the cameras at different positions in the storage area, judging the corresponding relation among the target detection frames, marking the target detection frames extracted by the different cameras and corresponding to the same material, personnel or equipment as the same identification number, and updating the corresponding prediction frames iteratively generated by the Kalman filtering unit; and the three-dimensional positioning unit is internally stored with a storage model obtained by scanning in advance and used for calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to the pixel positions of the target detection frame of the material target, the personnel target and the equipment target at different camera shooting angles.
Optionally, when the hungarian algorithm unit identifies the association between the target detection frame and the current frame prediction frame, if only the target detection frame exists and the current frame prediction frame matching the target detection frame does not exist, the material storage security management system marks a new identification number for the material target, the personnel target or the equipment target corresponding to the target detection frame, and adds a new track corresponding to the identification number; if the target detection frame is matched with the current frame prediction frame, continuously updating the target detection frame through a Kalman filtering unit to generate a prediction frame used for matching the next frame target detection frame and prediction frames corresponding to different camera shooting visual angle ranges; and if the target detection frame matched with the current frame prediction frame does not exist, deleting the track under the identification number of the corresponding current frame prediction frame.
Optionally, in any of the above plant leaf stomata individual behavior detection and analysis methods, the security check module is further provided with an action pattern recognition unit, which is used for calling different action recognition models according to a material target, a personnel target, and an equipment target to determine a movement state and an operation action of each target detection frame; the action recognition model is obtained in advance according to target action training.
Optionally, the material warehousing security management system as described in any one of the above, wherein the reaction terminal includes alarm devices pre-installed at different positions in the warehousing area, and a portable interactive device configured to the warehousing area patrol personnel, and configured to give an alarm prompt and correspondingly shoot and upload video images of targets exceeding the authority or not meeting the operation specification at corresponding positions in the warehousing area.
Optionally, the material warehousing security management system as described in any one of the above, wherein the warehouse control unit includes control units of door locks or transfer devices pre-installed at different positions in the warehousing area, and is configured to trigger, according to the judgment of the security check module on the operation of each target, a corresponding warehouse control unit arranged near the target to open a door lock passage permission to pass through or trigger a corresponding transfer device to move to a position where the target is located to receive and transfer a corresponding material when the target does not exceed an operation permission and meets an operation specification.
Meanwhile, in order to achieve the purpose, the application also provides a material storage security management method, which comprises the following steps: the method comprises the steps of firstly, receiving video images of a storage area shot by cameras arranged at different positions of the storage area; secondly, identifying, detecting and tracking the goods and materials target, the personnel target and the equipment target which are moved out of the image shelf area frame by frame according to the video image of the storage area shot by each camera, detecting and tracking the personnel target and determining the identity of the personnel target; thirdly, calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to pixel positions of the material target, the personnel target and the equipment target at different camera shooting angles; and fourthly, determining the moving state and the operation action of the warehouse model according to the specific coordinate positions of the material target, the personnel target and the equipment target relative to the warehouse model, triggering a reaction terminal arranged near the target to give an alarm prompt and store the tracking record of each camera when the operation exceeds the authority or does not accord with the operation specification, and triggering a corresponding warehouse control unit arranged near the target to pass when the operation does not exceed the authority and accords with the operation specification.
Optionally, in the second step of the method for managing security of material storage, a pre-trained yolov3 model is used to identify the detection targets contained in each frame of the video image, and correspondingly, target detection frames corresponding to the materials, the personnel or the equipment are extracted, and identification numbers corresponding to the target detection frames are marked; then, iteratively generating a frame prediction frame corresponding to each target detection frame through a Kalman filtering unit according to the coordinate position of each target detection frame in the previous frame and the target detection frames corresponding to the same material, personnel or equipment and shot by cameras at different positions in the storage area; and finally, respectively identifying the association condition before the target detection frame in each frame according to the matching degree of the pixel coordinate positions between the target detection frame and the prediction frame of the frame through a Hungarian algorithm, marking the mutually associated target detection frames in each frame as having the same identification number, and acquiring the tracks of different target detection frames through the identification numbers.
Optionally, the material storage security management method as described in any one of the above is characterized in that in the third step, modeling is performed according to a storage model obtained by scanning in advance, and then specific coordinate positions of a material target, a person target, and an equipment target corresponding to pixel positions of a target detection frame of a material target, a person target, and an equipment target at different camera shooting angles are determined according to the range of viewing angles shot by cameras at different positions in a storage area, where the material target, the person target, and the equipment target correspond to the storage model.
Advantageous effects
The application provides a goods and materials storage security protection management system, it is through mutually independent goods and materials monitoring module, personnel monitoring module and equipment monitoring module respectively independent realization to the discernment of goods and materials target, personnel target and equipment target in the storage area, detect and trail. Furthermore, the pixel moving track of the target detection frame can be converted into the corresponding three-dimensional coordinates through the storage model, so that the moving states and the operation actions of materials, personnel and equipment can be accurately extracted. From this, this application can surpass the authority or be not conform to the operating specification through security protection check-up module corresponding operation and trigger the warning suggestion in order to ensure warehouse safety to through the running state of comprehensive judgement personnel, equipment, goods and materials, further utilize the test function of model, when detecting and to accord with the authority requirement and operating specification, trigger warehouse control unit linkage when the operation of no safe risk and let pass in order to improve storage management transfer efficiency.
Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the application.
Drawings
The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application and not limit the application. In the drawings:
FIG. 1 is a schematic block diagram of a material warehousing security management system according to the present application;
FIG. 2 is a schematic diagram of warehouse images identified by the material warehousing security management system of the present application;
fig. 3 is a schematic diagram of an action pattern recognition technique adopted by the material warehousing security management system of the present application.
Detailed Description
In order to make the purpose and technical solutions of the embodiments of the present application clearer, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings of the embodiments of the present application. It should be apparent that the described embodiments are only some of the embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the described embodiments of the application without any inventive step, are within the scope of protection of the application.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The meaning of "and/or" as used herein is intended to include both the individual components or both.
The meaning of "inside and outside" in this application means that, with respect to the warehousing system itself, the direction from the edge of the warehouse image taken by the system to the center of the warehouse image is inside, and vice versa; and not as a specific limitation on the mechanism of the device of the present application.
The meaning of "left and right" in this application means that when the user is facing the warehouse images shot by the system of this application, the left side of the user is left, and the right side of the user is right, not the specific limitation of the mechanism of the device of this application.
The term "connected" as used herein may mean either a direct connection between components or an indirect connection between components via other components.
The meaning of "up and down" in this application means that when the user is facing the warehouse images captured by the system of this application, the direction from the bottom ground portion of the warehouse images to the top ceiling portion is up, otherwise it is down, and not specifically limited to the mechanism of the device of this application.
Fig. 1 is a material warehousing security management system according to the present application, which includes:
the cameras are arranged at different positions of the storage area and are used for shooting video images of the storage area at different viewing angles;
the material monitoring module is used for identifying, detecting and tracking a material target moved out of the image shelf area according to the storage area video images shot by the cameras;
the personnel monitoring module is used for identifying personnel targets in the storage area according to the video images of the storage area shot by the cameras, detecting and tracking the personnel targets and determining the identities of the personnel targets;
the equipment monitoring module is used for identifying, detecting and tracking equipment targets at channel positions among the shelf areas according to the video images of the storage areas shot by the cameras;
the storage model is connected with the material monitoring module, the personnel monitoring module and the equipment monitoring module and used for calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to the pixel positions of the material target, the personnel target and the equipment target at different camera shooting angles;
and the security and protection verification module is connected with the storage model and used for determining the moving state and the operation action of the storage model according to the specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model, triggering a reaction terminal arranged near the target to give an alarm prompt and store the tracking record of each camera when the operation exceeds the authority or does not accord with the operation specification, and triggering a corresponding storage control unit arranged near the target to pass when the operation does not exceed the authority and accords with the operation specification.
In the working process, the video images of the storage area are shot by using the cameras arranged at different positions of the storage area; then, according to the video images of the storage area shot by each camera, identifying, detecting and tracking the material target, the personnel target and the equipment target which are moved out of the image shelf area frame by frame, detecting and tracking the personnel target, and determining the identity of the personnel target; according to pixel positions of the material target, the personnel target and the equipment target under different camera shooting angles, specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model are calculated, and accordingly, the moving state and the operation action of the equipment target are determined according to the specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model, so that when the operation exceeds the authority or does not accord with the operation specification, a reaction terminal arranged near the target is triggered to give an alarm prompt and store the tracking record of each camera, and when the operation does not exceed the authority and accords with the operation specification, a corresponding storage control unit arranged near the target is triggered to pass.
In practical use, the camera can be generally arranged at the corner position of the passageway in the storage area so as to cover more passageway space, thereby detecting the movement of personnel and equipment in the storage area as comprehensively as possible. Referring to fig. 2, after the video images of the warehousing area are obtained, the system of the present application may extract a plurality of frames in the video at certain intervals according to the frame sequence of the video images to perform target extraction.
In the process of target extraction, detection targets such as materials, personnel and equipment contained in each frame of a video image can be identified by the material monitoring module, the personnel monitoring module and the equipment monitoring module respectively and independently utilizing yolov3 models trained by the modules in advance, target detection frames corresponding to the materials, the personnel and the equipment are correspondingly extracted, and identification numbers corresponding to the target detection frames are marked.
The goods and materials monitoring module can set goods and materials in the bin on the goods shelf into a reverse training set through selection of the training set, so that only the goods and materials extracted from the bin are screened out in the monitoring process, the goods and materials in the bin are removed, subsequent tracking operation is not performed, and the calculated amount of a subsequent tracking algorithm is reduced. The personnel monitoring module can be embedded into the existing face recognition algorithm, the personnel identity is confirmed by extracting the face characteristics and the body shape characteristics, the operation authority of the personnel is further inquired, and the goods taking behavior exceeding the authority is alarmed and correspondingly supervised. The equipment monitoring module can be embedded into an OCR character recognition module to acquire equipment number information, track the actual running state of each equipment and alarm to prompt related maintenance personnel to overhaul when detecting that the running track of the equipment has larger deviation with the running path set in the system.
In the specific process of tracking each target detection frame, the system of the application generally adopts the mode shown in fig. 3, and generates a frame prediction frame corresponding to each target detection frame through iteration of a kalman filtering unit according to the coordinate position of each target detection frame in the previous frame and the target detection frames corresponding to the same material, personnel or equipment, which are shot by cameras at different positions in a storage area; and then respectively identifying the association condition before the target detection frame in each frame according to the matching degree of the pixel coordinate positions between the target detection frame and the frame prediction frame by Hungary algorithm, marking the mutually associated target detection frames in each frame as having the same identification number so as to track different target detection frames through the identification numbers, and obtaining the motion trail of each target relative to the shooting position of the camera through the pixel coordinate position change condition of the target detection frames with the same number in the video.
In the specific operation process, due to the fact that the target detection frames of different cameras under different shooting visual angles need to be checked to judge the corresponding relation among the target detection frames under different visual angles, modeling needs to be carried out on the storage space in advance, the corresponding relation of each pixel point in the shot image of each camera relative to the three-dimensional coordinate position in the model is built according to the installation position and the shooting angle of each camera, and the area range of the target detection frame is mapped to the three-dimensional coordinate position in the storage model through the coordinate mapping matrix so that the corresponding relation among the target detection frames can be determined.
Therefore, in the process of tracking each target, the method firstly identifies the detection targets such as materials, personnel or equipment contained in each frame of the video image by using a pre-trained yolov3 model through a target detection unit, correspondingly extracts target detection frames corresponding to the materials, the personnel or the equipment, and marks identification numbers corresponding to the target detection frames;
then, generating a frame prediction frame according to each target detection frame in the previous frame through a Kalman filtering unit;
respectively identifying the association condition before the target detection frame in each frame according to the matching degree of the pixel coordinate positions between the target detection frame and the prediction frame of the frame by a Hungarian algorithm unit, and marking the mutually associated target detection frames in each frame as having the same identification number; in the matching process, target detection frames shot by different cameras can be converted into the visual angle ranges of other cameras through coordinate mapping matrixes according to the visual angle ranges shot by the cameras at different positions in the storage area respectively so as to judge the corresponding relation between the target detection frames shot by the different cameras, and the target detection frames extracted by the different cameras and corresponding to the same material, personnel or equipment are marked as the same identification number so as to update the corresponding prediction frames generated by the Kalman filtering unit in an iterative manner. For example, camera A and camera B set up the both ends of same passageway in the storage space respectively. When the matching degree between the target detection frames extracted by the two cameras is confirmed, the target detection frames extracted from the images shot by the camera A are converted into three-dimensional coordinate positions in the warehouse model through the coordinate mapping matrix corresponding to the camera A, then the target detection frames extracted from the images shot by the camera B are converted into the three-dimensional coordinate positions in the warehouse model through the coordinate mapping matrix corresponding to the camera B, the overlapping relation between the three-dimensional coordinate positions corresponding to the target detection frames in the two cameras is compared, when the coordinates of the two cameras are basically overlapped or the central points of the two three-dimensional coordinate positions are close to each other, the two target detection frames extracted from the two cameras are considered to correspond to the same target, and therefore the two target detection frames are marked to have the same identification number.
Specifically, when the association condition between the target detection frame and the frame prediction frame is identified through the Hungarian algorithm unit, according to the matching degree of the coordinate positions between the target detection frame and the frame prediction frame, when only the target detection frame exists and the frame prediction frame matched with the target detection frame does not exist, an identification number is singly added for marking a material target, a personnel target or an equipment target corresponding to the target detection frame, and a track corresponding to the identification number is newly added; or when the target detection frame is matched with the frame prediction frame, continuously updating the target detection frame through the Kalman filtering unit to generate a prediction frame used for matching the next frame target detection frame and prediction frames corresponding to different camera shooting visual angle ranges; or when there is no target detection frame matching with the frame prediction frame, the tracking is considered to be finished, and the track under the identification number corresponding to the frame prediction frame can be deleted.
The correlation condition between the target detection frame collected by different cameras and the corresponding frame prediction frame can be converted into a three-dimensional coordinate position in the warehouse model through a coordinate mapping matrix matched with the shooting visual angle of the cameras, so that matching is carried out through the proximity degree of the three-dimensional coordinate position between different detection frames in the warehouse model.
Meanwhile, the three-dimensional positioning unit can calculate specific coordinate positions of the material target, the personnel target and the equipment target relative to the warehousing model according to the warehousing model which is stored in the three-dimensional positioning unit and is obtained by scanning the warehousing space in advance and the target detection frame pixel positions of the material target, the personnel target and the equipment target at different camera shooting angles. So as to realize the tracking of the moving state and the motion trail of each target.
Therefore, according to the method and the device, the security check module can be triggered according to the three-dimensional coordinate position change condition of each target confirmed by the three-dimensional positioning unit in the storage space, and the action recognition models corresponding to different targets are called according to the material target, the personnel target and the equipment target through the internal pre-trained action pattern recognition unit, so that the movement state and the operation action of each target detection frame are determined through each action recognition model. Generally, the motion recognition model corresponding to each type of target can be obtained in advance according to the target motion training. For example, for a material target, a video image in the material flow storage process can be collected, the material to be identified is marked, and the video image is used as a training set to train a deep learning algorithm model, so that various states of goods taking out, turning, transferring and the like can be identified according to the moving state of a goods target detection frame in an actual video image. Aiming at the personnel target, the video image in the personnel moving process can be collected, the personnel needing to be identified and the positions of the hands and the feet thereof are marked, and the detection frames of the personnel and the positions of the hands and the feet thereof are used as a training set to train the deep learning algorithm model, so that various states such as manual goods taking, goods inspection, equipment operation, goods handling and the like can be identified according to the moving state of the personnel target detection frame in the actual video image. The action of the personnel is compared with the authority information of the personnel confirmed through face recognition, the alarm devices pre-installed at different positions of the storage area can be triggered when the personnel takes goods from the non-authority area or operates equipment without authority control, the portable interaction equipment and other reaction terminals configured for the storage area inspection personnel carry out alarm prompt, and video images of targets exceeding the authority or not meeting the operation specification at corresponding positions in the storage area are correspondingly shot and uploaded.
Similarly, for the equipment target, the video image in the equipment moving process can be collected, the position of a part needing to be identified is marked, the detection frame of the equipment and the position of the operation device of the equipment is used as a training set to train the deep learning algorithm model, and therefore the deep learning model can identify various states such as goods taking, goods inspection, goods transferring and goods unloading according to the moving state of the equipment target detection frame in the actual video image. Various actions of the equipment operation mechanism are compared with authority information corresponding to the equipment number confirmed by OCR character recognition, so that the equipment can be used for taking goods from an unauthorized area or triggering reaction terminals pre-installed at different positions of a storage area to give an alarm prompt when unauthorized operation is executed, and video images of targets exceeding the authority or not meeting the operation specification at corresponding positions in the storage area are correspondingly shot and uploaded.
For the targets which are matched in authority and meet the requirements of the operation specification, the control units of the door locks or the transfer equipment which are pre-installed at different positions of the storage area can be used for triggering the corresponding warehouse control units arranged near the targets to open the door lock passing authority for equipment personnel to pass through when the targets do not exceed the operation authority and meet the operation specification, or correspondingly triggering the transfer equipment capable of butting the targets to move to the positions of the targets so as to quickly realize butting and receiving of corresponding materials needing to be transferred.
To sum up, the material storage security management system that this application provided has following advantage compared with current system:
the system can establish a warehouse material dynamic network by using the Internet of things technology, integrate warehouse area monitoring resources, establish a comprehensive security management platform, implement a 'person-vehicle-single' lean management mode, associate and bind monitoring information with personnel information and service information, establish a relation map of security data and services, acquire and identify state information of the warehouse area in real time, realize all-round sensing and whole-process monitoring of a security system of a key area, and ensure the integrity and traceability of warehouse service data.
2-the system of the application can utilize the oblique photography technology to establish a three-dimensional point source, coordinate positioning is carried out by combining a real object ID, and a three-dimensional library area channel real scene three-dimensional model is drawn; the method comprises the steps of automatically acquiring material information by utilizing an image recognition technology, establishing a material object resource library for stored materials by combining a big data technology, realizing dynamic tracking, query and processing of material state information by differentiated data screening, establishing an interconnection and intercommunication mechanism with a service management system, avoiding asymmetry of material management information process and guaranteeing material management safety.
3-the system of the application can also utilize machine vision technology, accurately capture image information in an operation range in real time, and deeply excavate data meaning in the image information and behavior actions of analysts based on recognition analysis technology, and send out warning reminding information and accurately track and record through a built-in voice module of the safety helmet aiming at abnormal actions and irregular action early warning management and control, thereby reducing safety risks and improving the intelligent management and operation level of the warehouse.
4-the system of the application can greatly improve the material checking efficiency, and the checking time is reduced from the original 65.6 hours to 5.1 hours; the security management is changed from control solution to analysis prevention, and the occurrence of security events is avoided. In addition, the system can also accurately record the monitoring information of the business links and the operation places, and ensure the traceability of the security data of the warehouse; security data are integrated and analyzed, the problem of the process is optimized and improved, and an auxiliary decision is provided for material business. Therefore, the system can effectively enlarge the supervision range of managers and reduce the human resource cost and the warehousing management cost; and (4) the intelligent security control platform is popularized in internal units by combining the national network company warehouse security construction standard, and product income is created.
The above are merely embodiments of the present application, and the description is specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for those skilled in the art, without departing from the concept of the present application, several variations and modifications can be made, which are within the protection scope of the present application.
Claims (9)
1. The utility model provides a material storage security protection management system which characterized in that includes:
the cameras are arranged at a plurality of positions of the storage area and are used for shooting video images of the storage area;
the material monitoring module is used for identifying, detecting and tracking a material target moved out of the image shelf area according to the video images of the storage area shot by each camera;
the personnel monitoring module is used for identifying personnel targets in the storage area according to the video images of the storage area shot by the cameras, detecting and tracking the personnel targets and determining the identities of the personnel targets;
the equipment monitoring module is used for identifying, detecting and tracking equipment targets at channel positions among the shelf areas according to the video images of the storage areas shot by the cameras;
the storage model is connected with the material monitoring module, the personnel monitoring module and the equipment monitoring module and used for calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to the pixel positions of the material target, the personnel target and the equipment target at different camera shooting angles;
and the security and protection verification module is connected with the storage model and used for determining the moving state and the operation action of the storage model according to the specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model, triggering a reaction terminal arranged near the target to give an alarm prompt and store the tracking record of each camera when the operation exceeds the authority or does not accord with the operation specification, and triggering a corresponding storage control unit arranged near the target to pass when the operation does not exceed the authority and accords with the operation specification.
2. The material warehousing security management system of claim 1, wherein each of the material monitoring modules, the personnel monitoring module and the equipment monitoring module is independently provided with the following operation units for identifying, detecting and tracking a monitored target thereof:
the target detection unit is used for identifying detection targets contained in each frame of the video image by utilizing a pre-trained yolov3 model, correspondingly extracting target detection frames corresponding to materials, personnel or equipment, and marking identification numbers corresponding to the target detection frames;
a Kalman filtering unit for generating a frame prediction frame according to each target detection frame in the previous frame;
the Hungarian algorithm unit respectively identifies the association condition before the target detection frame in each frame according to the matching degree of the pixel coordinate positions between the target detection frame and the prediction frame of the frame, and marks the mutually associated target detection frames in each frame as having the same identification number;
the matching track unit is used for respectively remapping target detection frames shot by different cameras into the visual angle ranges of other cameras according to the visual angle ranges shot by the cameras at different positions in the storage area, judging the corresponding relation among the target detection frames, marking the target detection frames extracted by the different cameras and corresponding to the same material, personnel or equipment as the same identification number, and updating the corresponding prediction frames iteratively generated by the Kalman filtering unit;
and the three-dimensional positioning unit is internally stored with a storage model obtained by scanning in advance and used for calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to the pixel positions of the target detection frame of the material target, the personnel target and the equipment target at different camera shooting angles.
3. The material warehousing security management system of claim 2, wherein the association between the target detection box and the frame prediction box is identified by the Hungarian algorithm unit,
if only the target detection frame exists but the frame prediction frame matched with the target detection frame does not exist, an identification number is added for a material target, a personnel target or an equipment target corresponding to the target detection frame, and a track corresponding to the identification number is added;
if the target detection frame is matched with the current frame prediction frame, continuously updating the target detection frame through a Kalman filtering unit to generate a prediction frame used for matching the next frame target detection frame and prediction frames corresponding to different camera shooting visual angle ranges;
and if the target detection frame matched with the current frame prediction frame does not exist, deleting the track under the identification number of the corresponding current frame prediction frame.
4. The plant leaf stomata individual behavior detection and analysis method according to claim 2, characterized in that the security check module is provided with an action pattern recognition unit for calling different action recognition models according to material targets, personnel targets and equipment targets to determine the moving state and operation action of each target detection frame;
the action recognition model is obtained in advance according to target action training.
5. The material warehousing security management system of claims 1-4, wherein the reaction terminal comprises alarm devices pre-installed at different locations of a warehousing area, and portable interaction equipment configured for a warehousing area inspector to give alarm prompts and correspondingly shoot video images of objects exceeding the authority or not meeting the operating specifications at corresponding locations in the warehousing area.
6. The material warehousing security management system of claims 1-4, wherein the warehouse control unit comprises control units of door locks or transfer equipment pre-installed at different positions of a warehousing area, and is used for triggering corresponding warehouse control units arranged near the targets to open door lock passing authorities to pass through or triggering corresponding transfer equipment to move to the position of the target to receive and transfer corresponding materials when the targets do not exceed the operation authorities and meet the operation specifications according to the judgment of the security check module on the operation of each target.
7. A material storage security management method is characterized by comprising the following steps:
the method comprises the steps of firstly, receiving video images of a storage area shot by cameras arranged at different positions of the storage area;
secondly, identifying, detecting and tracking the material target, the personnel target and the equipment target moved out of the image shelf area frame by frame according to the video images of the storage area shot by each camera, detecting and tracking the personnel target and determining the identity of the personnel target;
thirdly, calculating specific coordinate positions of the material target, the personnel target and the equipment target relative to the storage model according to pixel positions of the material target, the personnel target and the equipment target at different camera shooting angles;
and fourthly, determining the moving state and the operation action of the warehouse model according to the specific coordinate positions of the material target, the personnel target and the equipment target relative to the warehouse model, triggering a reaction terminal arranged near the target to give an alarm prompt and store the tracking record of each camera when the operation exceeds the authority or does not accord with the operation specification, and triggering a corresponding warehouse control unit arranged near the target to pass when the operation does not exceed the authority and accords with the operation specification.
8. The material warehousing security management method of claim 7, wherein in the second step, the pre-trained yolov3 model is used to identify the detection targets contained in each frame of the video image, the target detection frames corresponding to the materials, personnel or equipment are extracted correspondingly, and the identification numbers corresponding to the target detection frames are marked;
then, iteratively generating a frame prediction frame corresponding to each target detection frame through a Kalman filtering unit according to the coordinate position of each target detection frame in the previous frame and the target detection frames corresponding to the same material, personnel or equipment and shot by cameras at different positions in the storage area;
and finally, respectively identifying the association condition before the target detection frame in each frame according to the matching degree of the pixel coordinate positions between the target detection frame and the prediction frame of the frame through a Hungarian algorithm, marking the mutually associated target detection frames in each frame as having the same identification number, and acquiring the tracks of different target detection frames through the identification numbers.
9. The material storage security management method according to claim 8, wherein in the third step, modeling is performed according to a storage model obtained by scanning in advance, and then specific coordinate positions of the material target, the personnel target and the equipment target corresponding to the pixel positions of the target detection frames of the material target, the personnel target and the equipment target at different camera shooting angles are determined according to the range of the visual angles shot by the cameras at different positions in the storage area, relative to the storage model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211102853.7A CN115604424A (en) | 2022-09-09 | 2022-09-09 | Material storage security protection management system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211102853.7A CN115604424A (en) | 2022-09-09 | 2022-09-09 | Material storage security protection management system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115604424A true CN115604424A (en) | 2023-01-13 |
Family
ID=84843863
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211102853.7A Pending CN115604424A (en) | 2022-09-09 | 2022-09-09 | Material storage security protection management system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115604424A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117291503A (en) * | 2023-09-28 | 2023-12-26 | 佳康捷科技(江苏)有限公司 | Intelligent warehouse management method and device and electronic equipment |
| CN117557968A (en) * | 2024-01-11 | 2024-02-13 | 深圳市明心数智科技有限公司 | Monitoring method, monitoring device, storage medium and computer equipment |
| CN117557201A (en) * | 2024-01-12 | 2024-02-13 | 国网山东省电力公司菏泽供电公司 | Intelligent warehouse safety management system and method based on artificial intelligence |
| CN117854211A (en) * | 2024-03-07 | 2024-04-09 | 南京奥看信息科技有限公司 | Target object identification method and device based on intelligent vision |
-
2022
- 2022-09-09 CN CN202211102853.7A patent/CN115604424A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117291503A (en) * | 2023-09-28 | 2023-12-26 | 佳康捷科技(江苏)有限公司 | Intelligent warehouse management method and device and electronic equipment |
| CN117557968A (en) * | 2024-01-11 | 2024-02-13 | 深圳市明心数智科技有限公司 | Monitoring method, monitoring device, storage medium and computer equipment |
| CN117557968B (en) * | 2024-01-11 | 2024-04-30 | 深圳市明心数智科技有限公司 | Monitoring method, monitoring device, storage medium and computer equipment |
| CN117557201A (en) * | 2024-01-12 | 2024-02-13 | 国网山东省电力公司菏泽供电公司 | Intelligent warehouse safety management system and method based on artificial intelligence |
| CN117557201B (en) * | 2024-01-12 | 2024-04-12 | 国网山东省电力公司菏泽供电公司 | Intelligent warehouse safety management system and method based on artificial intelligence |
| CN117854211A (en) * | 2024-03-07 | 2024-04-09 | 南京奥看信息科技有限公司 | Target object identification method and device based on intelligent vision |
| CN117854211B (en) * | 2024-03-07 | 2024-05-28 | 南京奥看信息科技有限公司 | Target object identification method and device based on intelligent vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115604424A (en) | Material storage security protection management system | |
| CN106412501B (en) | A kind of the construction safety behavior intelligent monitor system and its monitoring method of video | |
| CN108062349B (en) | Video monitoring method and system based on video structured data and deep learning | |
| CN109711370B (en) | Data fusion method based on WIFI detection and face clustering | |
| CN108319926A (en) | A kind of the safety cap wearing detecting system and detection method of building-site | |
| CN110619277A (en) | Multi-community intelligent deployment and control method and system | |
| CN111353338B (en) | Energy efficiency improvement method based on business hall video monitoring | |
| CN108805008A (en) | A kind of community's vehicle security system based on deep learning | |
| CN112132041A (en) | Community patrol analysis method and system based on computer vision | |
| CN110728252A (en) | Face detection method applied to regional personnel motion trail monitoring | |
| CN112183162A (en) | Face automatic registration and recognition system and method in monitoring scene | |
| CN110674761A (en) | Regional behavior early warning method and system | |
| CN112132048A (en) | Community patrol analysis method and system based on computer vision | |
| Patil et al. | Suspicious movement detection and tracking based on color histogram | |
| CN109086682A (en) | A kind of intelligent video black smoke vehicle detection method based on multi-feature fusion | |
| CN115100724A (en) | Multi-camera personnel state identification method | |
| US20180068172A1 (en) | Method of surveillance using a multi-sensor system | |
| CN116798176A (en) | Data management system based on big data and intelligent security | |
| CN111767776A (en) | Abnormal license plate selection method and device | |
| CN115035668A (en) | Community security system based on video monitoring | |
| CN110276379A (en) | A kind of the condition of a disaster information rapid extracting method based on video image analysis | |
| CN105095891A (en) | Human face capturing method, device and system | |
| CN113044694A (en) | Construction site elevator people counting system and method based on deep neural network | |
| CN111723725A (en) | Multi-dimensional analysis system based on video AI | |
| CN111553264A (en) | Campus non-safety behavior detection and early warning method suitable for primary and secondary school students |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |