Warehouse location monitoring method and device, computer equipment and storage medium

Abstract

The application relates to a warehouse storage location monitoring method and device, computer equipment and a storage medium. The method comprises the following steps: first, acquiring video data of a warehouse storage location area, and acquiring a target image based on the video data. Thus, based on a deep learning image recognition technology, the target image can be detected by a category detection model, and objects such as humans, vehicles and goods in the warehouse can be accurately and effectively recognized. Then, the state of the objects in each storage location is detected, and a detection result is obtained. The detection result is shared to a warehouse scheduling system in real time, so that real-time monitoring of the warehouse storage location is realized, and the accuracy of warehouse storage location recognition is greatly improved. In addition, the detection result in the application is shared to the warehouse scheduling system of the user in real time, which provides the enterprise with storage location safety information and assists the unmanned operation of intelligent equipment such as a patrol robot.

Description

Technical Field

[0001] This application relates to the field of warehouse location management technology, and in particular to a method, apparatus, computer equipment, and storage medium for monitoring warehouse locations. Background Technology

[0002] With the development of warehouse location management technology, the scenarios in warehouses are becoming increasingly complex, involving a mix of forklifts, staff, and goods. Laser sensors are often used to monitor warehouse locations in order to achieve warehouse location management.

[0003] However, the method using laser sensors can only monitor whether there are goods in the warehouse location, but cannot distinguish the type of the object being detected, which leads to an increase in false detection events and low accuracy in identifying warehouse locations. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, device, computer equipment, and storage medium for monitoring warehouse locations to address the aforementioned technical problems.

[0005] A method for monitoring warehouse locations, the method comprising:

[0006] The system acquires video data of the warehouse storage area and obtains target images corresponding to the storage area based on the video data. The warehouse storage area includes the storage area itself and the surrounding area. A category detection model is used to detect the target image and determine the category of each object appearing in the image. This category includes at least one of the following: human, vehicle, and goods. Based on the category of each object, the system detects the status of objects in each storage location and obtains detection results. These detection results include at least one of the following: whether a human has entered the warehouse storage area, vehicle status information, and storage location inventory information. The detection results are then sent to the warehouse scheduling system so that the system can monitor the warehouse storage area.

[0007] In one embodiment, acquiring video data of a warehouse storage area and obtaining a target image corresponding to the warehouse storage area based on the video data includes:

[0008] Acquire video data of the warehouse storage area; decode the video data to obtain a decoded image corresponding to the video data; calibrate the decoded image to obtain a calibrated image; downsample the calibrated image to obtain a target image corresponding to the warehouse storage area.

[0009] In one embodiment, the target image is detected based on a category detection model to determine the category to which each object appearing in the target image belongs, including:

[0010] Obtain a trained category detection model; based on the category detection model, extract features from the target image to obtain image features corresponding to the target image; determine the category of each object appearing in the target image based on the image features.

[0011] In one embodiment, based on the category to which each object belongs, the state of objects in each storage location is detected to obtain detection results, including:

[0012] When the category is a vehicle, determine whether the vehicle has entered the warehouse storage area, and issue an alarm when the vehicle enters the warehouse storage area; when the category is goods, perform state filtering on the area image representing the storage area in the video data, and determine the storage information based on the result of the state filtering; when the category is a human, determine whether the human has entered the warehouse storage area, and issue an alarm when the human enters the warehouse storage area.

[0013] In one embodiment, when the category is goods, state filtering is performed on the regional image representing the storage location area in the video data, and the storage location inventory information is determined based on the result of the state filtering, including:

[0014] When goods are identified as being present in the storage location, it is determined whether the goods are in an upright position. If the goods are in an upright position, multiple state filters are performed on the image representing the storage location in the video data to obtain multiple first state filter results corresponding to the number of state filters. It is then determined whether there are vehicles around the storage location. If multiple consecutive identical first state filter results are obtained, vehicles are present around the storage location, and the distance between the storage location and the vehicles is greater than a distance threshold, or if multiple consecutive identical first state filter results are obtained and no vehicles are present around the storage location, then the usage information of the storage location is determined, and the type of goods is determined through the goods information corresponding to the goods. The usage information of the storage location and the type of goods are used as the storage location inventory information of the storage location.

[0015] In one embodiment, when multiple and consecutive identical first state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is less than or equal to a distance threshold, the determination of the usage information of the storage area is stopped.

[0016] In one embodiment, when the storage area is determined to be empty, multiple state filters are performed on the area image representing the storage area in the video data to obtain multiple second state filter results corresponding to the number of state filters, and it is determined whether there are vehicles around the storage area; when multiple and consecutive identical second state filter results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is greater than a distance threshold, or when multiple and consecutive identical first state filter results are obtained, and no vehicles exist around the storage area, the usage information of the storage area is determined, and the usage information is used as the storage inventory information of the storage area.

[0017] A warehouse location monitoring device, the device comprising:

[0018] The acquisition module is used to acquire video data of the warehouse storage area and obtain the target image corresponding to the warehouse storage area based on the video data; the warehouse storage area includes the storage area and the area surrounding the storage area.

[0019] The determination module is used to detect the target image based on the category detection model and determine the category to which each object in the target image belongs. The category includes at least one of humans, vehicles, and goods.

[0020] The detection module is used to detect the status of objects in each storage location based on the category to which each object belongs, and obtain the detection results; the detection results include at least one of the following: whether a human has entered the warehouse storage area, vehicle status information, and storage location inventory information;

[0021] The sending module is used to send the detection result to the warehouse scheduling system so that the warehouse scheduling system can monitor the warehouse storage area.

[0022] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement a method for monitoring warehouse locations as described above.

[0023] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for monitoring warehouse locations as described above.

[0024] The aforementioned warehouse location monitoring method, device, computer equipment, and storage medium first acquire video data of the warehouse location area and then obtain target images corresponding to that area based on the video data. This allows for the detection of target images using deep learning image recognition technology and a category detection model, accurately and effectively identifying objects such as humans, vehicles, and goods within the warehouse. Furthermore, the status of objects in each location can be detected, and the detection results are shared in real-time with the warehouse scheduling system, thus achieving real-time monitoring of warehouse locations and significantly improving the accuracy of warehouse location identification. In addition, the inspection results detected in this application will be shared in real-time with the user's warehouse scheduling system, providing enterprises with warehouse location security information and assisting in the unmanned operation of intelligent equipment such as inspection robots. Attached Figure Description

[0025] Figure 1 This is an application environment diagram of a warehouse location monitoring method in one embodiment;

[0026] Figure 2 This is a flowchart illustrating a warehouse location monitoring method in one embodiment;

[0027] Figure 3 This is a usage scenario diagram of a warehouse location monitoring method in one embodiment;

[0028] Figure 4 This is a flowchart illustrating the steps for acquiring a target image in one embodiment;

[0029] Figure 5 This is a flowchart illustrating the steps for obtaining detection results in one embodiment;

[0030] Figure 6 This is a flowchart illustrating the steps for determining warehouse location inventory information when the category is goods, as shown in one embodiment.

[0031] Figure 7 This is a flowchart illustrating a method for detecting the state of objects in a warehouse storage area in one embodiment.

[0032] Figure 8 This is a flowchart illustrating the initial debugging process of a warehouse location monitoring method in one embodiment;

[0033] Figure 9 This is a flowchart illustrating a warehouse location monitoring method in another embodiment;

[0034] Figure 10 This is a data flow diagram of a warehouse location monitoring method in one embodiment;

[0035] Figure 11 This is a structural block diagram of a warehouse location monitoring device in one embodiment;

[0036] Figure 12 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0038] The warehouse location monitoring method provided in this application can be applied to, for example... Figure 1 In the application environment shown, camera device 102 communicates with computer device 104 via a network. Camera device 102 captures video of the warehouse storage area, obtaining corresponding video data, and transmits it to computer device 104. Computer device 104 acquires the video data and, based on the video data, obtains the target image corresponding to the warehouse storage area. Computer device 104 detects the target image based on a category detection model, determining the category of each object appearing in the target image. Based on the category of each object, computer device 104 detects the state of objects in each storage location, obtaining detection results. Computer device 104 sends the detection results to the warehouse scheduling system, enabling the warehouse scheduling system to monitor the warehouse storage locations. Camera device 102 can be, but is not limited to, various video acquisition devices, such as high-definition cameras, vision sensors, and mobile phones with camera functions. Computer device 104 can be a terminal or a server. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be a standalone server or a server cluster composed of multiple servers.

[0039] In one embodiment, such as Figure 2 As shown, a method for monitoring warehouse locations is provided, and this method is applied to... Figure 1 Taking a computer device as an example, the explanation includes the following steps:

[0040] Step 202: Obtain video data of the warehouse storage area and obtain a target image corresponding to the warehouse storage area based on the video data; the warehouse storage area includes the storage area and the area surrounding the storage area.

[0041] The storage area can be a warehouse storage area, and the area surrounding the storage area can be a loading platform area. The camera equipment can use a small number of low-cost visual sensors.

[0042] Specifically, the camera equipment performs video surveillance on the warehouse storage area to obtain video data of the warehouse storage area; the camera equipment sends the video data to the computer equipment through a switch, and the computer equipment receives the video data of the warehouse storage area; the computer equipment obtains the target image corresponding to the warehouse storage area based on the video data.

[0043] In one embodiment, the camera device and the computer device transmit video streams (video data transmission) through a switch. The camera device and the switch can transmit data streams through a wired network, and the switch and the computer device can transmit data streams through a wired network or mobile communication technology (such as 5G, 5th Generation Mobile Communication Technology).

[0044] In one embodiment, video data of warehouse locations is acquired in real time by cameras deployed in the storage area and loading platform. This video data is transmitted to a switch via a wired network. The monitoring program on the computer continuously acquires the video stream from the switch and uses this video stream to obtain target images corresponding to the warehouse locations. The cameras can be infrared-enabled, enabling real-time detection even in the absence of lighting, providing information support for intelligent equipment operating at night and reducing operating costs. This method is highly suitable for various flat-panel warehousing and cargo transfer scenarios in warehousing logistics, manufacturing, aviation, and other fields.

[0045] Step 204: Detect the target image based on the category detection model and determine the category to which each object in the target image belongs. The category includes at least one of human, vehicle, and cargo.

[0046] Specifically, the computer device inputs the acquired target image into a category detection model for detection. Based on the objects appearing in the target image, each object is classified according to its category to obtain the detection result. For example, the computer device inputs the target image into a category detection model, which classifies the objects appearing in the target image according to the categories of humans, vehicles, and goods to obtain the category of the objects appearing in the target image.

[0047] Step 206: Based on the category to which each object belongs, detect the status of the objects in each storage location and obtain the detection results; the detection results include at least one of the following: whether a human has entered the warehouse storage area, vehicle status information, and storage location inventory information.

[0048] The inventory information for warehouse locations includes whether the warehouse location is being used and the types of goods in the warehouse location.

[0049] Specifically, the computer equipment, based on the status assessment module, detects objects in each storage location according to a detection method corresponding to the category, and obtains detection results corresponding to the status of objects in each storage location. For example, the computer equipment uses the status assessment module to detect objects in each storage location, and obtains detection results corresponding to the category, such as the status of personnel in the storage location area, the status of vehicles, whether there are goods in the storage location, and the type of goods in the storage location. This enables intelligent sorting based on the type of goods.

[0050] Step 208: Send the detection result to the warehouse scheduling system so that the warehouse scheduling system can monitor the warehouse location area.

[0051] The warehouse scheduling system is the RCS system (Robots Control System), also known as the central control scheduling system, which is mainly used for robot scheduling.

[0052] Specifically, the computer equipment sends the detection results to the RCS system, which then schedules the warehouse storage area based on these results. For example, the computer equipment and the RCS can connect via Wi-Fi or 5G, and the detection results can be sent to the RCS system via HTTP and JSON-RPC to instruct the warehouse's unmanned forklifts to move.

[0053] The aforementioned warehouse location monitoring method first acquires video data of the warehouse location area and then obtains target images corresponding to the warehouse location area based on this video data. Deep learning image recognition technology is then used to detect the target images through a category detection model, accurately and effectively identifying objects such as people, vehicles, and goods in the warehouse. Furthermore, the status of objects in each location can be detected, and detection results are obtained. Finally, these detection results are shared in real-time with the warehouse scheduling system, thereby achieving real-time monitoring of warehouse locations and significantly improving the accuracy of warehouse location identification.

[0054] In one alternative embodiment, such as Figure 3As shown, the camera equipment acquires video data from the temporary storage area and the loading / unloading vehicles on the platform, and sends the video data to the switch via a wired network. The switch can then send the video data to the computer equipment via a wired network or 5G technology. The monitoring program on the computer equipment continuously acquires the video stream. Based on this video stream, the computer equipment acquires target images corresponding to the temporary storage area and the loading / unloading vehicles on the platform. Based on these target images, the computer equipment uses a category detection model to obtain the category of each object in the target image. Based on the category of each object, the computer equipment obtains the detection result corresponding to the category according to the detection method corresponding to the category. The computer equipment can connect to the RCS system via 5G technology or Wi-Fi and send the detection results to the RCS system via HTTP and JSON_RPC to instruct the actions of the unmanned forklift in the warehouse. The RCS system and the unmanned forklift can connect via 5G technology or Wi-Fi, thereby realizing real-time monitoring of warehouse locations and improving the accuracy of warehouse location identification.

[0055] In one embodiment, such as Figure 4 As shown, video data of the warehouse storage area is acquired, and a target image corresponding to the warehouse storage area is obtained based on the video data, including:

[0056] Step 402: Obtain video data of the warehouse storage area.

[0057] Specifically, the monitoring program on the computer equipment continuously acquires video streams of the warehouse storage area via a switch. For example, the storage location detection camera inputs the video data corresponding to the warehouse storage area into the image processing module of the computer equipment, which is used for image acquisition and preprocessing.

[0058] Step 404: Decode the video data to obtain a decoded image corresponding to the video data.

[0059] Specifically, after the computer device inputs video data into the image processing module, it decodes the video data to obtain a decoded image corresponding to the video data.

[0060] Step 406: The decoded image is calibrated to obtain a calibrated image.

[0061] Specifically, different camera devices have different models, which causes distortion in the decoded image. The computer device adjusts the parameters of the decoded image to correct the distorted area and obtain a calibrated image.

[0062] Step 408: Downsample the calibration image to obtain the target image corresponding to the warehouse storage area.

[0063] Specifically, the computer device downsamples the calibration image to obtain a target image corresponding to the warehouse location area, thereby reducing the computational load on the image, and the computer device saves the video data corresponding to the warehouse location area.

[0064] In this embodiment, the computer device acquires video data of the warehouse storage area, and based on the video data, performs decoding processing to obtain a decoded image, realizing the process of converting video into an image; then, by calibrating the decoded image, a calibrated image is obtained, thereby solving the problem of image distortion caused by different camera devices; then, the computer device obtains the target image by downsampling the calibrated image, thereby reducing the amount of computation required for the target image.

[0065] In one embodiment, detecting the target image based on a category detection model and determining the category to which each object in the target image belongs includes: acquiring a trained category detection model; extracting features from the target image based on the category detection model to obtain image features corresponding to the target image; and determining the category to which each object in the target image belongs based on the image features.

[0066] Specifically, after acquiring the target image corresponding to the warehouse storage area, the computer device loads a trained category detection model and inputs the target image into the model for feature extraction. Once the model completes its forward computation, it obtains the image features corresponding to the target image. Based on these features, the model classifies the objects appearing in the target image to determine their respective categories. For example, the computer device inputs the acquired target image into a deep learning object detection and classification module, loads the trained classification model, performs forward computation, obtains features corresponding to people, goods, and vehicles in the target image, and organizes these features to determine the category of each object in the target image.

[0067] In this embodiment, the computer device obtains image features corresponding to the target image through a trained category detection model, and determines the category of each object appearing in the target image based on these image features. Therefore, deep learning technology can reduce dependence on the environment (such as site, surroundings, lighting, etc.) and can effectively identify personnel, goods, and vehicles in warehouse storage areas, thereby achieving intelligent classification and improving the accuracy of warehouse location monitoring and identification.

[0068] In one embodiment, such as Figure 5 As shown, based on the category to which each object belongs, the status of objects in each storage location is detected, and the detection results are obtained, including:

[0069] Step 502: When the category is vehicle, determine whether the vehicle has entered the warehouse storage area, and issue an alarm message when the vehicle enters the warehouse storage area.

[0070] Specifically, the computer equipment detects warehouse storage areas based on the camera equipment location allocation configuration table. When the category is "vehicle," if the computer equipment detects a vehicle entering the warehouse storage area, it issues an alarm. This alarm can be triggered via indicator lights and horns deployed in the warehouse storage area, sent to the computer equipment's display interface, or sent to the RCS system. For example, when the category is "vehicle," the computer equipment detects whether a vehicle has entered the warehouse storage area. If a vehicle enters, the horns and indicator lights in the warehouse will issue a warning through sound and light, respectively. Simultaneously, an alarm text will be sent to the computer equipment's display interface and the RCS system, continuing until the vehicle leaves the warehouse storage area.

[0071] Step 504: When the category is goods, perform state filtering on the regional image representing the storage location area in the video data, and determine the storage location inventory information based on the result of the state filtering.

[0072] Specifically, the computer equipment detects each storage location based on the camera equipment storage location allocation configuration table. When the category is goods, the computer equipment sets the number of filtering times and performs multiple state filtering processes on the regional image of the standard storage location area in the video data to obtain the state filtering result corresponding to the number of filtering times. Based on the state filtering processing result, the storage location inventory information is determined.

[0073] In one embodiment, the computer device obtains the camera device storage location allocation configuration table through the state assessment module, detects each storage location in the warehouse storage location area, and when the category is goods, performs corresponding state filtering on each frame of the region image representing the storage location area in the video data, obtains the state filtering result corresponding to the frame number, and compares the state filtering result corresponding to the previous frame of the region image with the state filtering result corresponding to the current frame of the region image based on the state filtering result corresponding to each frame of the region image, obtains multiple comparison results, and determines the storage location inventory information based on the comparison results.

[0074] In one embodiment, the computer device obtains the camera device storage location allocation configuration table through the state assessment module, detects each storage location in the warehouse storage location area, and when the category is goods, performs multiple state filtering processes on the current frame region image based on the region image corresponding to the current frame representing the storage location area in the video data, obtains the state filtering result corresponding to the number of filtering times, compares the previous state filtering result and the current state filtering result in the current frame region image to obtain multiple comparison results, and determines the storage location inventory information based on the comparison results.

[0075] Step 506: When the category is human, determine whether the human has entered the warehouse storage area, and issue an alarm message when the human enters the warehouse storage area.

[0076] Specifically, when the category is "human," the computer equipment determines whether a human has entered the warehouse storage area by detecting the area. When a human is detected entering the warehouse storage area, the computer equipment issues an alarm. This alarm can be triggered via indicator lights and speakers deployed in the warehouse storage area, or it can be sent to the computer equipment's display interface or to the RCS system. For example, when the category is "human," the computer equipment detects whether anyone has entered the warehouse storage area. If a human enters the warehouse storage area or the surrounding area, the speakers and indicator lights in the warehouse will issue a warning through sound and light, respectively. Simultaneously, an alarm message will be sent to the computer equipment's display interface and then to the RCS system, continuing until the human leaves the warehouse storage area.

[0077] In this embodiment, when the category is a vehicle, it is determined whether the vehicle has entered the warehouse storage area, and an alarm is issued when the vehicle enters the warehouse storage area. When the category is goods, state filtering is performed on the regional image representing the storage area in the video data, and the storage information of the storage location is determined based on the result of the state filtering. When the category is a human, it is determined whether the human has entered the warehouse storage area, and an alarm is sent when the human enters the warehouse storage area. Therefore, by performing detection according to the detection method corresponding to the category, timely safety monitoring of personnel and vehicles can be carried out, and timely acquisition of storage information of each storage location in the warehouse storage area can be achieved, thereby improving the accuracy of storage location monitoring and identification.

[0078] In one embodiment, such as Figure 6 As shown, when the category is goods, state filtering is performed on the regional image representing the storage location area in the video data, and the storage location inventory information is determined based on the result of the state filtering, including:

[0079] Step 602: When goods are found in the storage area, determine whether the goods are in the correct position.

[0080] Specifically, the computer equipment detects the storage location area. When the storage location area determines that there are goods in the storage location, it determines whether the goods are in an upright position based on the positional relationship between the storage location and the goods. The positional relationship between the storage location and the goods refers to whether the entire goods are within the storage location area.

[0081] In one embodiment, when the computer device determines that the goods are not in the correct position, the loudspeakers and indicator lights in the warehouse will issue a warning through sound and light, respectively. At the same time, an alarm text will be sent to the display interface of the computer device, and the alarm information will be sent to the RCS system.

[0082] Step 604: When the goods are in the upright position, perform multiple state filtering based on the area image representing the storage location in the video data to obtain multiple first state filtering results corresponding to the number of state filtering times, and determine whether there are vehicles around the storage location area.

[0083] Specifically, based on the camera equipment storage location allocation configuration table, each storage location is detected. When the category is goods and the goods are in the correct position, the computer equipment sets the number of filtering times and performs multiple state filtering based on the regional image of the standard storage location in the video data to obtain multiple first state filtering results; the computer equipment determines whether there are vehicles around the storage location area.

[0084] Step 606: When multiple and consecutive identical first state filtering results are obtained, there are vehicles around the storage area, and the distance between the storage area and the vehicles is greater than the distance threshold, or when multiple and consecutive identical first state filtering results are obtained, and there are no vehicles around the storage area, the usage information of the storage area is determined, and the type of goods is determined by the goods information corresponding to the goods.

[0085] Specifically, the computer device acquires multiple first-state filtering results. When it acquires the same first-state filtering results that are consistent with the number of filtering iterations, detects that there is a vehicle in the storage area, and the distance between the storage area and the vehicle is greater than a distance threshold, or when it acquires multiple and consecutive first-state filtering results and detects that there are no vehicles around the storage area, the computer device determines the storage space occupancy status in the storage area based on the first-state filtering results, and determines the type of goods by acquiring goods information.

[0086] In one embodiment, the state assessment module of the computer device obtains the camera device storage location allocation configuration table, detects each storage location in the warehouse storage location area, and when the category is goods and there are goods in the storage location, sets the number of state filtering operations, such as performing state filtering 20 times within 1 minute. Based on the representation of the storage location area in the video data, at least 20 frames of area images are obtained, and corresponding state filtering is performed on each frame of area image to obtain the corresponding first state filtering result. The computer device also detects whether there are vehicles around the storage location area.

[0087] For example, when the first-state filtering result shows 20 consecutive instances of "goods available" and there are no vehicles around the storage location, the filtering result corresponding to the previous frame's regional image is compared with the filtering result corresponding to the current frame's regional image. If all 20 comparison results show "goods available," the storage location is determined to be in a "goods available" state, and the storage location is displayed in red on the computer device's display interface, indicating that it has goods available. Alternatively, when the first-state filtering result shows 20 consecutive instances of "goods available," there are vehicles around the storage location, and the distance between the vehicle and the storage location is greater than a distance threshold (e.g., the distance threshold is 2m, and the distance between the vehicle and the storage location exceeds 2m), the filtering result corresponding to the previous frame's regional image is compared with the filtering result corresponding to the current frame's regional image. If all 20 comparison results show "goods available," the storage location is determined to be in a "goods available" state, and the storage location is displayed in red on the computer device's display interface, indicating that it has goods available. If the filtering count is not reached, the computer device increases the filtering count until 20 consecutive identical filtering results appear. The computer device uses the goods information to determine the type of goods.

[0088] In one embodiment, the state assessment module of the computer device obtains the camera device storage location allocation configuration table, detects each storage location in the warehouse storage location area, and when the category is goods and there are goods in the storage location, sets the number of state filtering operations, such as 20 state filtering operations within 1 minute. Based on the area image corresponding to the current frame representing the storage location area in the video data, the current frame area image is subjected to at least 20 state filtering operations to obtain the corresponding first state filtering result, and the computer device detects whether there are vehicles around the storage location area.

[0089] For example, when the first-state filtering result shows 20 consecutive instances of "goods available" and there are no vehicles around the storage location, the previous filtering result in the current frame image is compared with the current filtering result. If all 20 comparison results show "goods available," the storage location is determined to be in a "goods available" state, and the location is displayed in red on the computer device's display interface, indicating that the goods are available. Alternatively, when the first-state filtering result shows 20 consecutive instances of "goods available," there are vehicles around the storage location, and the distance between the vehicle and the storage location is greater than a distance threshold (e.g., the distance threshold is 2m, and the distance between the vehicle and the storage location exceeds 2m), the previous filtering result in the current frame image is compared with the current filtering result. If all 20 comparison results show "goods available," the storage location is determined to be in a "goods available" state, and the location is displayed in red on the computer device's display interface, indicating that the goods are available. If the required number of filtering iterations is not reached, the computer device increases the number of filtering iterations until 20 consecutive identical filtering results are obtained. The computer device uses goods information to determine the type of goods.

[0090] Step 608: Use the usage information of the storage location area and the type of goods as the storage location inventory information of the storage location area.

[0091] Specifically, the computer equipment uses the usage information of the storage location area and the types of goods in the storage location as the storage location inventory information. This computer equipment saves the storage location inventory information and outputs it to the RCS system and the computer equipment's display module. For example, the computer equipment sends the usage status of the storage location and the types of goods in the storage location as the storage location inventory information to the RCS system via HTTP and JSON_RPC to instruct the actions of the warehouse's unmanned forklift. The RCS system and the unmanned forklift can be connected via 5G technology or Wi-Fi. When the computer equipment's interface program issues an image acquisition command, the image is input to the computer equipment's display module via HTTP and Rpclib.

[0092] In this embodiment, when goods are confirmed to be present in a storage location, the computer device determines whether the goods are in an upright position to ensure their safe storage. When the goods are in an upright position, multiple state filters are performed on the image representing the storage location in the video data to obtain a first state filter result, and it is determined whether there are vehicles around the storage location. If multiple consecutive identical first state filter results are obtained, vehicles are present around the storage location, and the distance between the storage location and the vehicles is greater than a distance threshold; or if multiple consecutive identical first state filter results are obtained, and no vehicles are present around the storage location, then the usage information of the storage location is determined. The type of goods is then determined based on the goods information corresponding to the goods, and the usage information and the type of goods are used as the storage location inventory information. Therefore, not only can the occupancy status of each storage location in the storage area be obtained in a timely manner, but the type of goods stored in the storage location can also be obtained, thereby timely obtaining the inventory information of each storage location in the warehouse storage area and improving the accuracy of storage location monitoring and identification.

[0093] In one embodiment, when multiple consecutive identical first state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is less than or equal to a distance threshold, the determination of the usage information of the storage area is stopped.

[0094] Specifically, the computer device obtains multiple first-state filtering results. When it obtains the same first-state filtering results that are consistent with the number of filtering cycles, detects that there are vehicles around the storage area, and the distance between the storage area and the vehicles is less than or equal to the distance threshold, it stops determining the usage information of the storage area. Determining the usage information of the storage area means refreshing the occupancy status of the storage area.

[0095] In one embodiment, the state assessment module of the computer device obtains the storage location allocation configuration table of the camera device, detects each storage location in the warehouse storage area, and when the category is goods and there are goods in the storage location (i.e., a vehicle places goods in the storage location), the number of state filtering operations is set, such as 20 state filtering operations within 1 minute. Based on the representation of the storage location area in the video data, at least 20 frames of area images are obtained, and the corresponding state filtering is performed on each frame of area image to obtain the corresponding first state filtering result. Furthermore, the distance between the storage location area and the vehicle is less than or equal to a distance threshold (i.e., the vehicle has not yet left the area around the storage location where goods have been placed). This distance threshold can be set by the computer device, for example, the computer device sets the distance threshold to 2 meters.

[0096] When the automated forklift receives a storage instruction from the RCS system and enters the storage area, the instruction is for the forklift to store goods in an empty storage location, marked in blue. The computer display will then mark locations with goods in red and empty locations in blue. After the forklift delivers the goods to the storage location, the computer performs a status filter. If the first status filter result for that location shows "goods present" for the first 20 consecutive times, and the forklift is less than or equal to 2 meters from the location, the computer display stops refreshing for that location; it will still display the location as blue. Simultaneously, the warehouse's speakers and indicator lights will issue warnings via sound and light, and an alarm message will be sent to the computer display and the RCS system. This process continues until the distance between the storage location and the forklift exceeds 2 meters, at which point the computer will restart refreshing the location, changing the location's marker from blue to red.

[0097] In this embodiment, when multiple consecutive identical first-state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is less than or equal to a distance threshold, the determination of the usage information of the storage area is stopped. Therefore, the computer equipment can monitor the distance between the vehicle and the storage area in a timely manner, and thus provide timely feedback on the vehicle's operating status, thereby achieving safe monitoring of the vehicle.

[0098] In one embodiment, when the storage area is determined to be empty, multiple state filters are performed on the image representing the storage area in the video data to obtain multiple second state filter results corresponding to the number of state filters, and it is determined whether there are vehicles around the storage area; when multiple and consecutive identical second state filter results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is greater than a distance threshold, or when multiple and consecutive identical first state filter results are obtained, and no vehicles exist around the storage area, the usage information of the storage area is determined, and the usage information is used as the storage inventory information of the storage area.

[0099] Specifically, based on the camera equipment storage location allocation configuration table, each storage location is detected. When the storage location area is determined to be empty, the computer equipment sets the number of filtering iterations and performs multiple state filtering based on the area image of the standard storage location in the video data to obtain multiple second state filtering results. The computer equipment determines whether there are vehicles around the storage location area. The computer equipment obtains multiple second state filtering results. When the same second state filtering result with the same number of filtering iterations is obtained, the computer equipment detects the presence of vehicles around the storage location area, and the distance between the storage location area and the vehicles is greater than a distance threshold, or when multiple and consecutive first state filtering results are obtained, and the computer equipment detects that there are no vehicles around the storage location area, the storage location occupancy status in the storage location area is determined based on the second state filtering results, and this usage information is used as the storage location inventory information for that storage location area.

[0100] In one embodiment, the computer device obtains the camera device storage location allocation configuration table through the state assessment module, detects each storage location in the warehouse storage area, and when the storage location area is determined to be empty, sets the number of filtering times, such as performing state filtering 20 times within 1 minute. Based on the representation of the storage location area in the video data, at least 20 frames of area images are obtained, and corresponding state filtering is performed on each frame of area image to obtain the corresponding second state filtering result. The computer device also detects whether there are vehicles around the storage location area.

[0101] For example, when the second-state filtering result shows "out of stock" for the first time 20 consecutive times and there are no vehicles around the storage location, the filtering result corresponding to the previous frame's regional image is compared with the filtering result corresponding to the current frame's regional image. If all 20 comparison results show "out of stock," the storage location is determined to be out of stock, and the storage location is displayed in blue on the computer device's display interface, indicating that it is out of stock. Alternatively, when the first-state filtering result shows "out of stock" for the first time 20 consecutive times, there are vehicles around the storage location, and the distance between the vehicle and the storage location is greater than a distance threshold (e.g., the distance threshold is 2m, and the distance between the vehicle and the storage location exceeds 2m), the filtering result corresponding to the previous frame's regional image is compared with the filtering result corresponding to the current frame's regional image. If all 20 comparison results show "out of stock," the storage location is determined to be out of stock, and the storage location is displayed in blue on the computer device's display interface, indicating that it is out of stock. If the required number of filtering iterations is not reached, the computer device increases the number of filtering iterations until 20 consecutive identical filtering results are obtained.

[0102] In one embodiment, the computer device obtains the camera device storage location allocation configuration table through the state assessment module, detects each storage location in the warehouse storage area, and when the storage location area is determined to be empty, sets the number of filtering times, such as 20 state filtering times within 1 minute. Based on the area image corresponding to the current frame representing the storage location area in the video data, the current frame area image is subjected to at least 20 state filtering times to obtain the corresponding second state filtering result. The computer device also detects whether there are vehicles around the storage location area.

[0103] For example, when the second-state filtering result shows "out of stock" for the first time 20 consecutive times and there are no vehicles around the storage location, the previous filtering result in the current frame area image is compared with the current filtering result. If all 20 comparison results show "out of stock," the storage location is determined to be out of stock, and the storage location is displayed in blue on the computer device's display interface, indicating that it is out of stock. Alternatively, when the first-state filtering result shows "out of stock" for the first time 20 consecutive times, there are vehicles around the storage location area, and the distance between the vehicle and the storage location area is greater than a distance threshold (e.g., the distance threshold is 2m, and the distance between the vehicle and the storage location area exceeds 2m), the previous filtering result in the current frame area image is compared with the current filtering result. If all 20 comparison results show "out of stock," the storage location is determined to be out of stock, and the storage location is displayed in blue on the computer device's display interface, indicating that it is out of stock. If the required number of filtering iterations is not reached, the computer device increases the number of filtering iterations until 20 consecutive identical filtering results are obtained.

[0104] In this embodiment, when a storage location is determined to be empty, multiple state filters are performed on the regional image representing the storage location in the video data to obtain a second state filter result, and it is determined whether there are vehicles around the storage location. When multiple consecutive identical second state filter results are obtained, vehicles exist around the storage location, and the distance between the storage location and the vehicles is greater than a distance threshold; or when multiple consecutive identical first state filter results are obtained, and no vehicles exist around the storage location, the usage information of the storage location is determined, and this usage information is used as the storage location's inventory information. Therefore, the inventory information of each storage location in the warehouse storage area can be obtained in a timely manner to indicate the storage locations available for unmanned forklift warehouses.

[0105] In one embodiment, when the unmanned forklift receives a pickup instruction from the RCS system and drives out of the storage area, the instruction is for the unmanned forklift to retrieve goods from a storage location marked in red. After the unmanned forklift retrieves the goods from the storage location, the storage location is filtered for status. When the first status filter result of the storage location appears consecutively 20 times for the first time, indicating that the goods are present, and the distance between the unmanned forklift and the storage location is less than or equal to 2 meters, the computer device's display interface stops refreshing the storage location. That is, the computer device's display interface still shows the storage location as red, and the speakers and indicator lights in the warehouse provide warnings through sound and light, respectively. At the same time, an alarm text is sent to the computer device's display interface, and the alarm information is sent to the RCS system. This continues until the distance between the storage location and the unmanned forklift is greater than 2 meters, at which point the computer device restarts refreshing the storage location, that is, the storage location changes from red to blue.

[0106] In this embodiment, when multiple consecutive identical first-state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is less than or equal to a distance threshold, the determination of the usage information of the storage area is stopped. Therefore, the computer equipment can monitor the distance between the vehicle and the storage area in a timely manner, and thus provide timely feedback on the vehicle's operating status, thereby achieving safe monitoring of the vehicle.

[0107] To facilitate a clearer understanding of the object status detection scheme in the warehouse storage area, a detailed implementation example is provided below. The specific implementation scheme for detecting the object status in the warehouse storage area is as follows: Figure 7As shown, the monitoring system can be understood as computer equipment. The monitoring system performs status detection based on the category of objects within the warehouse storage area. First, it detects the warehouse storage area and its individual storage locations. When personnel enter the entire warehouse storage area, the warehouse's speakers and indicator lights issue warnings via sound and light, respectively. Simultaneously, an alarm message is sent to the monitoring system's display interface and to the RCS system, continuing until the person leaves the warehouse storage area. When detecting individual storage locations within the warehouse area, it uses the camera storage location allocation table to determine if there are vehicles or goods in each location. When a vehicle is detected, an alarm is triggered via the warehouse's speakers and indicator lights, an alarm message is sent to the monitoring system's display interface and to the RCS system, continuing until the vehicle leaves the storage location. When goods are in a storage location, it first checks if the goods are placed correctly. If they are misplaced, the monitoring system can issue an alarm via the display interface, speakers, and indicator lights, and send the alarm message to the RCS system to ensure the goods are safely stored. When a storage location contains goods and the goods are in an upright position, or when the storage location is empty, a filtering count is set, and the system checks whether the current filtering result matches the previous filtering result, whether the filtering count has been reached, and whether a vehicle appears during the filtering process. Specifically, if consecutive identical filtering results occur, the filtering count is reached, a vehicle is present around the storage location, and the distance between the storage location and the vehicle is greater than a distance threshold; or if consecutive identical filtering results occur, the filtering count is reached, and no vehicle is present around the storage location, the storage location status is changed, and the storage location occupancy status is recorded. If the storage location contains goods, the monitoring system determines the type of goods by obtaining the goods information in the storage location. Therefore, when a storage location contains goods, the monitoring system records the storage location occupancy status and goods type information and outputs them to the RCS system to instruct the warehouse unmanned forklift to operate; when a storage location is empty, the monitoring system records the storage location occupancy status and outputs it to the RCS system to instruct the warehouse unmanned forklift to operate. Therefore, the monitoring system can detect the status of each object according to its category, thereby timely obtaining the inventory information of each storage location in the warehouse storage area, thus improving the accuracy of storage location monitoring and identification.

[0108] To facilitate a clearer understanding of the technical solution of this application, a more detailed embodiment is provided below. When monitoring warehouse locations, the initial debugging of the monitoring system is performed first. The monitoring system can be understood as computer equipment, specifically as follows: Figure 8As shown, the process involves installing and adjusting the detection cameras (i.e., video equipment), setting up the monitoring server environment, calibrating the camera's intrinsic parameters, and using a tiling tool to divide the warehouse storage area. Detection data is then acquired, including data on a single person, multiple people, a single vehicle, multiple vehicles, vehicles at various angles and positions, manned forklifts, personnel near vehicles, single goods, multiple goods (i.e., goods of various types, spacing, angles, and positions), vehicles moving goods, vehicles next to goods, vehicles moving goods with personnel nearby, goods with personnel and vehicles nearby, people operating forklifts, and people standing on pallets (if pallets are present). The detection data undergoes data clearing, sample balancing, data labeling, and model training to obtain a trained category detection model. This model is then imported into the test program, and the running model is configured. Finally, the RCS system is connected to the monitoring system for integration testing. After the integration test is completed, the debugged monitoring system is used in applications such as... Figure 1 The application environment shown.

[0109] For warehouse location monitoring solutions, the specifics are as follows: Figure 9 As shown, the monitoring program in the monitoring system first continuously acquires video streams from the location detection cameras in the camera equipment. The computer equipment can be the monitoring system itself. The video streams are then input to the image acquisition and preprocessing module for processing to obtain the target image. This module decodes the input video to obtain a decoded image, adjusts the distortion in the decoded image, performs downsampling to reduce computation, and saves the original video stream. The target image is then input to the deep learning object detection and classification module for classification to obtain the category of each object appearing in the target image. This involves first loading the trained category detection model, and then inputting the target image... The data is fed into a category detection model for forward computation to obtain the category to which each object belongs. The category to which each object belongs is then input into the status assessment module, which uses the obtained status and type of goods in the storage location, the status of personnel in the temporary storage area, and the status of vehicles in the temporary storage area as the status results. Finally, the status results are recorded in the monitoring system, and the monitoring system outputs the status results to the RCS system and the interface display system via HTTP+Json_rpc. If the interface program issues a command to acquire an image, the image is input into the display module of the monitoring system via HTTP+Rpclib, thereby realizing real-time monitoring of warehouse storage locations and improving the accuracy of warehouse storage location identification.

[0110] Because the interface display module (i.e., the interface system) transmits images and status information over the network, the interface program can be deployed on warehouse monitoring servers, customer servers, tablets, etc., making it highly portable. The display program of the interface display module allows users to click on the camera equipment overview map or enter a fixed camera identifier to select a specific camera image for preview. Furthermore, the interface display module can also be remotely controlled to complete some configured functions.

[0111] In the above embodiments, the flow of video data within the storage location monitoring interface (i.e., the interface display module) of the monitoring system, the deep learning storage location monitoring perception module of the monitoring system, and the RCS system can be specifically referred to. Figure 10 The bidirectional arrows in the diagram represent bidirectional data flow. The XML configuration table for storage locations corresponds to the camera storage location allocation configuration table, containing the coordinates, intrinsic and extrinsic parameters, and storage location information for each camera. The deep learning-based storage location monitoring and perception module corresponds to the monitoring system; the storage location monitoring interface corresponds to the display module output by the monitoring system. The deep learning-based storage location monitoring and perception module and the display interface communicate via Json_rpc and Rpclib. Json_rpc is a public interface that communicates with both the interface program and the RCS system, primarily transmitting the status of storage locations, people, and vehicles. Since Json_rpc cannot transmit images, RpclibMessage pack is used to transmit images to the interface program.

[0112] This document describes the flow of video data within the storage location monitoring interface (i.e., the interface display module), the deep learning storage location monitoring and perception module, and the RCS system of the monitoring system. The monitoring system can be understood as a computer device. First, the storage location monitoring interface sends the camera IDs (i.e., the camera devices) to the deep learning storage location monitoring and perception module. The deep learning storage location monitoring and perception module then transmits images back to the storage location monitoring interface. The interface program can control the perception module to perform calibration, data collection, and log recording, and can read and write XML storage location tables. Data flow between the deep learning storage location monitoring and perception module and the RCS system is bidirectional. The deep learning storage location monitoring and perception module sends status updates to the RCS system, and the RCS system can also query the locations of personnel and vehicles individually, or allow the deep learning storage location monitoring and perception module to remain unrefreshed or initialized.

[0113] It should be understood that, although Figure 2 , 4 The steps in flowcharts -7 and 9-10 are shown sequentially as indicated by the arrows; however, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order requirement for the execution of these steps, and they can be performed in other orders. Furthermore, Figure 2 , 4-7, at least some of the steps in 9-10 may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the steps or stages in other steps.

[0114] In one embodiment, such as Figure 11 As shown, a warehouse location monitoring device is provided, comprising: an acquisition module 1102, a determination module 1104, a detection module 1106, and a transmission module 1108, wherein:

[0115] The acquisition module 1102 is used to acquire video data of the warehouse storage area and obtain a target image corresponding to the warehouse storage area based on the video data; the warehouse storage area includes the storage area and the area surrounding the storage area.

[0116] The determination module 1104 is used to detect the target image based on the category detection model and determine the category to which each object appearing in the target image belongs. The category includes at least one of human, vehicle, and cargo.

[0117] The detection module 1106 is used to detect the status of objects in each storage location based on the category to which each object belongs, and obtain the detection result; the detection result includes at least one of the following: whether a human has entered the warehouse storage area, vehicle status information, and storage location inventory information.

[0118] The sending module 1108 is used to send the detection result to the warehouse scheduling system so that the warehouse scheduling system can monitor the warehouse storage area.

[0119] In one embodiment, the acquisition module 1102 is specifically used to acquire video data of the warehouse storage area; decode the video data to obtain a decoded image corresponding to the video data; calibrate the decoded image to obtain a calibrated image; and downsample the calibrated image to obtain a target image corresponding to the warehouse storage area.

[0120] In one embodiment, the determining module 1104 is specifically used to acquire a trained category detection model; based on the category detection model, perform feature extraction on the target image to obtain image features corresponding to the target image; and determine the category to which each object appearing in the target image belongs based on the image features.

[0121] In one embodiment, the detection module 1106 is specifically used to determine whether a vehicle has entered the warehouse storage area when the category is a vehicle, and to issue an alarm message when the vehicle enters the warehouse storage area; when the category is goods, to perform state filtering processing on the regional image representing the storage area in the video data, and to determine the storage information based on the result of the state filtering processing; and when the category is humans, to determine whether a human has entered the warehouse storage area, and to issue an alarm message when a human enters the warehouse storage area.

[0122] In one embodiment, the detection module 1106 is used to determine whether the goods are in a straight position when the storage area is found to contain goods; when the goods are in a straight position, it performs multiple state filtering based on the area image representing the storage location in the video data to obtain multiple first state filtering results corresponding to the number of state filtering times, and determines whether there are vehicles around the storage area; when multiple and consecutive identical first state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is greater than a distance threshold, or when multiple and consecutive identical first state filtering results are obtained, and no vehicles exist around the storage area, the usage information of the storage area is determined, and the type of goods is determined through the goods information corresponding to the goods; the usage information of the storage area and the type of goods are used as the storage inventory information of the storage area.

[0123] In one embodiment, the detection module 1106 is further configured to stop determining the usage information of the storage area when multiple and consecutive identical first state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is less than or equal to a distance threshold.

[0124] In one embodiment, the detection module 1106 is further configured to, when the storage area is determined to be free of goods, perform multiple state filtering operations on the area image representing the storage location in the video data to obtain multiple second state filtering results corresponding to the number of state filtering operations, and determine whether there are vehicles around the storage area; when multiple and consecutive identical second state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is greater than a distance threshold, or when multiple and consecutive identical first state filtering results are obtained, and no vehicles exist around the storage area, then determine the usage information of the storage area and use the usage information as the storage inventory information of the storage area.

[0125] Specific limitations regarding warehouse location monitoring devices can be found in the above description of warehouse location monitoring methods, and will not be repeated here. Each module in the aforementioned warehouse location monitoring device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0126] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 12 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores monitoring data for warehouse locations. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a method for monitoring warehouse locations.

[0127] Those skilled in the art will understand that Figure 12 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0128] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0129] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0130] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0131] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0132] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for monitoring warehouse locations, characterized in that, The method includes: Acquire video data of the warehouse storage area, and obtain a target image corresponding to the warehouse storage area based on the video data; the warehouse storage area includes the storage area and the area surrounding the storage area. The target image is detected based on a category detection model to determine the category to which each object in the target image belongs. The category includes at least one of human, vehicle, and cargo. When the category is goods, it is determined whether the goods are in a correct position based on the positional relationship between the storage location and the goods. The positional relationship refers to whether the entire goods are within the storage location area. When the goods are placed correctly, based on the regional images of each frame representing the storage location area in the video data, the corresponding state filtering is performed on the regional image of each frame to obtain the state filtering result corresponding to the frame number. Based on the state filtering result corresponding to each frame regional image, the state filtering result corresponding to the previous frame regional image is compared with the state filtering result corresponding to the current frame regional image. When multiple and consecutive identical first state filtering results are obtained, there are vehicles around the storage location area, and the distance between the storage location area and the vehicles is greater than the distance threshold, or when multiple and consecutive identical first state filtering results are obtained, and there are no vehicles around the storage location area, the storage location in the storage location area is determined to be in a state of having goods based on the first state filtering, and the type of goods is determined by the goods information corresponding to the goods. The usage information of the storage location area and the type of goods are used as the storage location inventory information of the storage location area; The inventory information of the storage location is sent to the warehouse scheduling system so that the warehouse scheduling system can monitor the storage location area.

2. The method according to claim 1, characterized in that, The step of acquiring video data of the warehouse storage area and obtaining a target image corresponding to the warehouse storage area based on the video data includes: Acquire video data of the warehouse storage area; The video data is decoded to obtain a decoded image corresponding to the video data; The decoded image is then calibrated to obtain a calibrated image; The calibration image is downsampled to obtain the target image corresponding to the warehouse storage area.

3. The method according to claim 1, characterized in that, The step of detecting the target image based on a category detection model and determining the category of each object appearing in the target image includes: Obtain the trained category detection model; Based on the category detection model, feature extraction is performed on the target image to obtain image features corresponding to the target image; The category to which each object in the target image belongs is determined based on the image features.

4. The method according to claim 1, characterized in that, The method further includes: When the category is a vehicle, determine whether the vehicle has entered the warehouse storage area, and issue an alarm message when the vehicle enters the warehouse storage area. When the category is human, determine whether the human has entered the warehouse storage area, and issue an alarm message when the human enters the warehouse storage area.

5. The method according to claim 1, characterized in that, The method further includes: When multiple consecutive identical first-state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is less than or equal to a distance threshold, the determination of the usage information of the storage area is stopped.

6. The method according to claim 1, characterized in that, The method further includes: When it is determined that there are no goods in the storage area, multiple state filters are performed on the area image representing the storage area in the video data to obtain multiple second state filter results corresponding to the number of state filters, and it is determined whether there are vehicles around the storage area. When multiple consecutive identical second-state filtering results are obtained, vehicles exist around the storage area, and the distance between the storage area and the vehicles is greater than a distance threshold, or when multiple consecutive identical first-state filtering results are obtained, and no vehicles exist around the storage area, the usage information of the storage area is determined, and the usage information is used as the storage inventory information of the storage area.

7. A monitoring device for warehouse storage locations, characterized in that, The device includes: The acquisition module is used to acquire video data of the warehouse storage area and obtain a target image corresponding to the warehouse storage area based on the video data; the warehouse storage area includes the storage area and the area surrounding the storage area. The determination module is used to detect the target image based on the category detection model and determine the category to which each object appearing in the target image belongs, wherein the category includes at least one of human, vehicle and cargo; The detection module is used to determine whether the goods are in an upright position based on the positional relationship between the storage location and the goods, where the positional relationship refers to whether the entire goods are within the storage location area. When the goods are in an upright position, based on the regional images representing the storage location area in the video data, the module performs corresponding state filtering on each frame's regional image to obtain the state filtering result corresponding to the frame number. Based on the state filtering results corresponding to each frame's regional image, the module compares the state filtering result corresponding to the previous frame's regional image with the state filtering result corresponding to the current frame's regional image. If multiple consecutive identical first state filtering results are obtained, vehicles exist around the storage location area, and the distance between the storage location area and the vehicles is greater than a distance threshold, or if multiple consecutive identical first state filtering results are obtained, and no vehicles exist around the storage location area, then the module determines that the storage location in the storage location area is in a stocked state based on the first state filtering. The module also determines the type of goods through the goods information corresponding to the goods. The usage information of the storage location area and the type of goods are used as the storage location inventory information of the storage location area. The sending module is used to send the inventory information of the storage location to the warehouse scheduling system so that the warehouse scheduling system can monitor the storage location area of ​​the warehouse.

8. The apparatus according to claim 7, characterized in that, The acquisition module is used to acquire video data of the warehouse storage area; decode the video data to obtain a decoded image corresponding to the video data; calibrate the decoded image to obtain a calibrated image; and downsample the calibrated image to obtain a target image corresponding to the warehouse storage area.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images