Solid waste monitoring method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a solid waste monitoring method and device, electronic equipment and a storage medium. Obtaining a target video stream of the to-be-monitored warehouse according to a preset analysis period; identifying a region of interest in the target video stream; solid waste is identified in the region of interest, and inventory parameters for the identified solid waste are determined. According to the embodiment of the invention, the accuracy of the warehouse-in and warehouse-out data of the solid waste is improved, and the requirement on computing resources is reduced.

Description

Technical Field

[0001] Embodiments of the present application relate to image processing technology, and more particularly to a solid waste monitoring method, device, electronic device, and storage medium. Background Art

[0002] Solid waste can cause certain harm to the ecological environment. In order to supervise solid waste, it is necessary to monitor the relevant data of solid waste in and out of the warehouse.

[0003] In the prior art, the transport vehicles in the solid waste storage warehouse are identified to monitor the in-and-out data of the solid waste.

[0004] However, transport vehicles are relatively free, which may cause data errors, and the transport data of combined vehicles is needed to determine the in and out data of solid waste. It is necessary to coordinate data from multiple parties and have high requirements for computing resources. Summary of the invention

[0005] The present application provides a solid waste monitoring method, device, electronic device and storage medium to improve the accuracy of solid waste in-and-out storage data and reduce the requirements for computing resources.

[0006] In a first aspect, an embodiment of the present application provides a solid waste monitoring method, the solid waste monitoring method comprising:

[0007] According to the preset analysis cycle, obtain the target video stream of the warehouse to be monitored;

[0008] Identify regions of interest in a target video stream;

[0009] Solid wastes in the area of ​​interest are identified and inventory parameters for the identified solid wastes are determined.

[0010] In a second aspect, an embodiment of the present application further provides a solid waste monitoring device, the solid waste monitoring device comprising:

[0011] A target video stream acquisition module is used to acquire the target video stream of the warehouse to be monitored according to a preset analysis cycle;

[0012] An area of ​​interest recognition module, used to recognize an area of ​​interest in a target video stream;

[0013] The inventory parameter determination module is used to identify solid waste in the area of ​​interest and determine inventory parameters of the identified solid waste.

[0014] In a third aspect, an embodiment of the present application further provides an electronic device, the electronic device comprising:

[0015] one or more processors;

[0016] A storage device for storing one or more programs;

[0017] When one or more programs are executed by one or more processors, the one or more processors implement any one of the solid waste monitoring methods provided in the embodiments of the present application.

[0018] In a fourth aspect, an embodiment of the present application further provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute any one of the solid waste monitoring methods provided in the embodiments of the present application.

[0019] The present application obtains the target video stream of the warehouse to be monitored according to the preset analysis cycle, and monitors the warehouse through the video, without the need for manual on-site verification, thereby reducing labor costs; identifies the area of ​​interest in the target video stream, reduces the monitoring area of ​​solid waste, improves the accuracy of subsequent monitoring, and reduces the requirements for computing resources; identifies the solid waste in the area of ​​interest, and determines the inventory parameters of the identified solid waste, and directly monitors the solid waste in the warehouse to be monitored through the target video stream, thereby reducing the interference of factors such as vehicles, improving the accuracy of inventory parameters, and does not need to analyze the behavior of the vehicle, nor does it need to coordinate other data, thereby reducing the requirements for computing resources. Therefore, through the technical solution of the present application, the problem that the transportation vehicles are relatively free, which may cause data errors, and the need to coordinate data from multiple parties, which has a high requirement for computing resources, is solved, and the accuracy of the inventory parameters of solid waste is improved, thereby improving the accuracy of the in-and-out data obtained according to the inventory parameters, and reducing the requirements for computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a flow chart of a solid waste monitoring method in Example 1 of the present application;

[0021] Figure 2a is a flow chart of a solid waste monitoring method in Example 2 of the present application;

[0022] Figure 2b This is a schematic diagram of solid waste packaging in Example 2 of the present application;

[0023] Figure 3a is a flow chart of a solid waste monitoring method in Embodiment 3 of the present application;

[0024] Figure 3b It is a schematic diagram of an overall area type in the third embodiment of the present application;

[0025] Figure 3c This is a schematic diagram of a boundary demarcation type in Embodiment 3 of the present application;

[0026] Figure 3d It is a schematic diagram of a physical boundary type in Embodiment 3 of the present application;

[0027] Figure 3e It is a schematic diagram of a manual calibration type in Embodiment 3 of the present application;

[0028] Figure 4 It is a structural schematic diagram of a solid waste monitoring device in Embodiment 4 of the present application;

[0029] Figure 5 It is a structural diagram of an electronic device in Embodiment 5 of the present application. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0031] It should be noted that the terms "first" and "second" etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] Embodiment 1

[0033] Figure 1 This is a flow chart of a solid waste monitoring method provided in Example 1 of the present application. This embodiment can be applied to monitoring inventory data of solid waste in a warehouse where solid waste is stored. The method can be executed by a solid waste monitoring device, which can be implemented using software and / or hardware and specifically configured in a monitoring platform of a solid waste warehouse.

[0034] See also Figure 1 The solid waste monitoring method shown in the figure specifically comprises the following steps:

[0035] S110. Obtain a target video stream of the warehouse to be monitored according to a preset analysis cycle.

[0036] The analysis cycle can be a pre-configured cycle for analyzing the target video stream. The process of solid waste in and out of the warehouse takes a certain amount of time, and the in and out of the warehouse does not necessarily occur in real time, so there is no need to analyze the target video stream in real time, and the limitation of computer resources may not meet the requirements of real-time analysis, so the analysis cycle can be set to reduce the demand for computing resources. Exemplarily, the analysis cycle can be determined by professional technicians based on factors such as the frequency of in and out of the warehouse and the computing power of the computer.

[0037] The warehouse to be monitored may be a warehouse that needs to be monitored, that is, a warehouse that stores solid waste. The target video stream may be a video stream that captures the solid waste storage area in the warehouse to be monitored. Exemplarily, a video stream list may be set, and the video stream list may store the captured content and acquisition address of the video stream, etc. The target video stream may be determined according to the video stream list, and the target video stream may be periodically acquired according to the address of the target video stream according to a preset analysis period.

[0038] S120: Identify a region of interest in the target video stream.

[0039] The region of interest may be an area where solid waste is placed in the warehouse to be monitored, which is used to improve the accuracy of subsequent solid waste identification. Region of interest identification is a commonly used technology in image processing and computer vision, which can extract specific areas from images or video frames for more detailed analysis and processing. For example, the region of interest in the target video stream can be identified by a deep learning algorithm.

[0040] S130, identifying solid waste in the area of ​​interest, and determining inventory parameters of the identified solid waste.

[0041] Solid waste can be solid waste generated during production. There may be some hazardous substances in solid waste, so it is necessary to monitor the inventory parameters of solid waste. For example, solid waste can be packaged in ton barrels, ton bags, ton boxes or special wooden boxes and placed in a special warehouse, that is, the warehouse to be monitored.

[0042] The inventory parameters may be related parameters describing the storage quantity of solid waste in the warehouse to be monitored, which are used for subsequent analysis of the storage and warehousing situation. Exemplarily, the inventory parameters may be the number, floor space, stacking height, etc., which are not specifically limited in this application. Exemplarily, the solid waste in the area of ​​interest may be identified through a target recognition algorithm, and the inventory parameters of the identified solid waste may be determined. For example, the target recognition algorithm may be a convolutional neural network, a support vector machine, or a random forest, which are not specifically limited in this application.

[0043] In order to prevent hazardous waste in solid waste from causing damage to the environment and ecology, according to the Law on the Prevention and Control of Environmental Pollution by Solid Waste and the management requirements of relevant departments for hazardous waste, it is necessary to timely fill in the ledger and report the subsequent information of the solid waste in and out of the warehouse parameters, and on this basis, use other management means to prevent and prevent illegal operations. In order to ensure that the solid waste monitoring work is in place, it is necessary to automatically analyze the inventory of hazardous waste in the warehouse through video monitoring and other means, and identify the increase or decrease in its quantity, so as to match the reported in and out of the warehouse ledger or joint form and other necessary management items, and make real-time predictions on suspected violations or untimely reporting, enhance off-site supervision capabilities, reduce the frequency of on-site law enforcement, and improve management efficiency.

[0044] In the prior art, solid waste is monitored through license plate recognition or forklift behavior recognition. The license plate recognition solution generally sets up video surveillance at the factory gate to recognize the license plates of vehicles entering and leaving the warehouse, and matches the recognition results with the warehouse entry and exit documents to achieve early prediction of abnormal documents. The forklift behavior recognition solution sets up monitoring probes at the main traffic arteries in the factory, such as warehouse entrances and exits, to capture the behavior of forklifts, determine whether the forklifts are loaded, and compare them with the current warehouse entry and exit ledger to detect abnormal entry and exit behavior.

[0045] However, there is a high risk of misjudgment when monitoring solid waste through license plate recognition or forklift behavior recognition. The license plate recognition solution is mainly used in the scenarios of joint orders leaving and arriving at the factory, and can be compared with offline actual operations. For example, scenarios such as outsourcing utilization and disposal. In these scenarios, the license plate and other information on the joint order are required to be completely accurate. Since the existing license plate recognition accuracy can reach more than 90%, the current operation effect is very good. However, it is not applicable to some self-utilization and disposal behaviors that do not enter or leave the factory area, or other situations that affect inventory changes. It has great limitations and can only achieve a few off-site supervision scenarios. There are two main problems with the forklift behavior recognition solution: on the one hand, the frequency of forklift movement behavior is high, and most of the cases are not related to the entry and exit of solid waste or joint orders. Analyzing it will cause a lot of meaningless operations interference, which will lead to a waste of computing power; on the other hand, the movement of forklifts is relatively free. At present, the judgment of the direction of forklift movement is not technically mature, which will cause a lot of misjudgments and interfere with the evaluation and analysis results. Many judgment results need to be further evaluated manually.

[0046] The technical solution of this embodiment obtains the target video stream of the warehouse to be monitored according to the preset analysis cycle, and monitors the warehouse through the video, without the need for manual on-site verification, thereby reducing labor costs; identifies the area of ​​interest in the target video stream, reduces the monitoring area of ​​solid waste, improves the accuracy of subsequent monitoring, and reduces the requirements for computing resources; identifies the solid waste in the area of ​​interest, and determines the inventory parameters of the identified solid waste, and directly monitors the solid waste in the warehouse to be monitored through the target video stream, thereby reducing the interference of factors such as vehicles, improving the accuracy of inventory parameters, and does not need to analyze the behavior of the vehicle, nor does it need to coordinate other data, thereby reducing the requirements for computing resources. Therefore, through the technical solution of this application, the problem that the transportation vehicles are relatively free, which may cause data errors, and the need to coordinate data from multiple parties, which has a high requirement for computing resources, is solved, and the accuracy of the inventory parameters of solid waste is improved, thereby improving the accuracy of the in-and-out data obtained according to the inventory parameters, and reducing the requirements for computing resources.

[0047] Embodiment 2

[0048] Figure 2a This is a flow chart of a solid waste monitoring method provided in Example 2 of the present application. The technical solution of this embodiment is further refined on the basis of the above technical solution.

[0049] Furthermore, "identifying solid waste in the area of ​​interest and determining inventory parameters of the identified solid waste" is refined as: "identifying solid waste in the area of ​​interest through a trained solid waste packaging recognition model; determining inventory parameters of the identified solid waste based on the area parameters of the warehouse to be monitored and the identified solid waste, the inventory parameters including at least one of quantity and floor area" to determine the inventory parameters of the identified solid waste.

[0050] See also Figure 2a A solid waste monitoring method is shown, comprising:

[0051] S210. Obtain the target video stream of the warehouse to be monitored according to a preset analysis cycle.

[0052] S220: Identify a region of interest in the target video stream.

[0053] S230: Identify solid waste in the area of ​​interest using the trained solid waste packaging recognition model.

[0054] The solid waste packaging recognition model may be an image recognition model for identifying solid waste packaging. Exemplarily, the solid waste packaging recognition model may be a large model. Optionally, the solid waste packaging recognition model may be a YOLO (professional term, a network model) model. Solid waste will be packaged using a fixed type of packaging. Figure 2b A schematic diagram of solid waste packaging. Figure 2b From left to right in the figure are ton barrels, ton bags, ton boxes and special wooden boxes. The solid waste packaging recognition model can be trained with solid waste packaging sample data to obtain a trained solid waste packaging recognition model, so that the trained solid waste packaging recognition model can accurately identify the solid waste packaging in the area of ​​interest. The area of ​​interest in the identified target video stream is input into the trained solid waste packaging recognition model, so that the solid waste in the area of ​​interest can be identified and the inventory parameters can be further obtained.

[0055] S240. Determine inventory parameters of the identified solid waste according to the area parameters of the warehouse to be monitored and the identified solid waste, where the inventory parameters include at least one of quantity and floor space ratio.

[0056] The area parameter of the warehouse to be monitored may be a relevant parameter for calculating the area of ​​the warehouse to be monitored. For example, the area parameter of the warehouse to be monitored may be the length and width of the warehouse to be monitored. In a warehouse to be monitored with a dedicated storage area, the area parameter may be the length and width of the dedicated storage area.

[0057] The inventory parameters include at least one of quantity and floor space ratio. The quantity may be the number of solid wastes identified. The floor space ratio may be the ratio of the floor space of the identified solid waste to the area of ​​the warehouse to be monitored. In a warehouse to be monitored with a dedicated storage area, the floor space ratio may be the ratio of the floor space of the identified solid waste to the area of ​​the dedicated storage area. After the trained solid waste packaging recognition model identifies the solid waste in the area of ​​interest, it may count the identified solid waste to obtain the quantity, determine the length and width of the area occupied by the identified solid waste, and multiply the length and width to obtain the solid waste floor space, and combine the area parameters of the warehouse to be monitored to obtain the floor space ratio.

[0058] In an optional embodiment, the solid waste packaging recognition model is trained in the following manner: labeling the packaging images of various types of solid waste to obtain a sample data set; building an initial solid waste packaging recognition model; training the initial solid waste packaging recognition model based on the sample data set to obtain a trained solid waste packaging recognition model.

[0059] The sample data set may be a data set including images of various types of solid waste packaging. For example, for each different type of solid waste packaging, about 500 images may be obtained and manually labeled with sample labels.

[0060] The initial solid waste packaging recognition model may be a network model that has not been trained with sample data. Exemplarily, the initial solid waste packaging recognition model may be a YOLO model. For example, relevant YOLO codes may be downloaded from GitHub (a professional term, a hosting platform for open source and private software projects), pre-trained weight files suitable for the sample data set may be downloaded, and relevant configuration work may be performed to build the initial solid waste packaging recognition model.

[0061] The sample data set can be divided into a training set and a test set according to a preset ratio. For example, the training set and the test set can be divided in a ratio of 8:2. Input the data of the training set into the initial solid waste packaging recognition model to train the model, and test the trained model through the test set. Running the training script generally may take 2 to 3 weeks. During this process, it is necessary to observe the changes in the loss function, the adjustment of the learning rate, and the performance on the validation set. If necessary, some hyperparameters need to be adjusted according to the progress of training. Evaluate the performance of the model on the test set, check parameters such as recall rate, precision rate, and mean average precision (mAP). If the performance does not meet the requirements, it is necessary to fine-tune the model or change the training strategy. When the requirements are met on the test set, a trained solid waste packaging recognition model is obtained.

[0062] By marking the packaging images of various types of solid waste, a sample data set is obtained, and a labeled sample data set is obtained, so that supervised training can be performed on the initial solid waste packaging recognition model; an initial solid waste packaging recognition model is built; the initial solid waste packaging recognition model is trained according to the sample data set to obtain a trained solid waste packaging recognition model. By supervised training the initial solid waste packaging recognition model, the accuracy of the trained solid waste packaging recognition model can be guaranteed, and the accuracy of subsequent identification of solid waste packaging can be improved.

[0063] The technical solution of this embodiment identifies solid waste in an area of ​​interest through a trained solid waste packaging recognition model. The solid waste in the area of ​​interest is directly identified without identifying the transport vehicle and with the cooperation of a corresponding database, thereby reducing the requirements for computer resources and improving the recognition accuracy through the trained solid waste packaging recognition model. According to the area parameters of the warehouse to be monitored and the identified solid waste, the inventory parameters of the identified solid waste are determined, and the inventory parameters include at least one of the quantity and the floor area. The inventory situation of the solid waste is comprehensively described through the quantity and the floor area, providing multiple dimensions of data support for subsequent inventory change analysis and avoiding accidental errors in single data.

[0064] Embodiment 3

[0065] Figure 3a This is a flow chart of a solid waste monitoring method provided in Example 3 of the present application. The technical solution of this embodiment is further refined on the basis of the above technical solution.

[0066] Furthermore, "identifying the region of interest in the target video stream" is refined as follows: "determining a target boundary recognition model from candidate boundary recognition models according to the storage boundary type of the warehouse to be monitored; identifying the region of interest in the target video stream according to the target boundary recognition model" to determine and identify the region of interest in the target video stream.

[0067] See also Figure 3a A solid waste monitoring method is shown, comprising:

[0068] S310. Obtain the target video stream of the warehouse to be monitored according to a preset analysis cycle.

[0069] S320: Determine a target boundary recognition model from candidate boundary recognition models according to the storage boundary type of the warehouse to be monitored.

[0070] The storage boundary type may be the type of the boundary for storing solid waste in the warehouse to be monitored. Exemplarily, the storage boundary type may include an overall area type, a physical boundary type, a demarcated boundary type, and an artificially demarcated type. According to the storage boundary type of the warehouse to be monitored, a target boundary recognition model may be determined from the candidate boundary recognition models to identify the region of interest in the target video stream.

[0071] Figure 3b It is a schematic diagram of an overall regional type; Figure 3c A schematic diagram for demarcating a type of boundary; Figure 3d is a schematic diagram of a physical boundary type; Figure 3e A schematic diagram of a manual calibration type. Figure 3e The red marked part is the manually marked region of interest.

[0072] In an optional embodiment, the candidate boundary recognition models include: an overall warehouse recognition model, a physically partitioned warehouse recognition model, a boundary line recognition model, and a borderless warehouse recognition model.

[0073] Different candidate boundary recognition models can be models for identifying different storage boundary types. In the overall area type scenario, the entire warehouse can generally be used to stack solid waste, leaving only some obvious boundary areas for aisles. In this case, the machine model can generally be trained to recognize boundaries to obtain an overall warehouse recognition model. The overall warehouse recognition model is used as the target boundary recognition model for the overall area type warehouse to be monitored.

[0074] Physical boundary type scenarios usually involve building walls and partitions in the warehouse, and some also add markings. The partitions divide the small areas into areas of interest for solid waste storage. In this case, the machine model can generally be trained to identify the boundaries and obtain the corresponding physical partition warehouse identification model. The physical partition warehouse identification model is used as the target boundary identification model for the physical boundary type warehouse to be monitored.

[0075] The boundary type demarcation scenario is generally performed by physically marking the warehouse with obvious colors (such as yellow) to divide the area of ​​interest for stacking hazardous waste packaging. In this case, the specific color is generally directly identified after color space conversion. Therefore, the boundary line recognition model identifies the specific color after color space conversion. The boundary line recognition model is used as the target boundary recognition model for the warehouse to be monitored for demarcating the boundary type.

[0076] In a warehouse with manual calibration type scenario, it is generally difficult to identify the stacking area through obvious markings. Therefore, manual marking can be performed first, and then recognition can be performed. Generally, the recognition result area can be directly provided to the running program as the area of ​​interest. The boundaryless warehouse recognition model is used as the target boundary recognition model for the warehouse to be monitored with manual calibration type.

[0077] The overall warehouse recognition model and the physical partition warehouse recognition model need to be trained by supervision. The training process includes data preparation, model architecture building and supervised training.

[0078] Data preparation includes preparing video frame photo data of similar warehouses, manually marking areas of interest, and performing image preprocessing.

[0079] Building a model architecture includes selecting a suitable model architecture. For example, the model architecture can select Fast R-CNN (a professional term, a network model architecture), and then design the model structure, including the input layer, convolution layer, pooling layer, fully connected layer, and output layer.

[0080] Training the model includes splitting the data, forming training sets, validation sets, and test sets, configuring training parameters, and conducting supervised training. During supervised training, weights are adjusted through backpropagation to minimize the loss function. At the end of each round, the performance of the trained model is evaluated using the validation set, and hyperparameters are adjusted as needed to complete the model training.

[0081] The candidate boundary recognition models include: overall warehouse recognition model, physical partition warehouse recognition model, boundary line recognition model and borderless warehouse recognition model. Different types of warehouses are recognized through different models to improve the accuracy of identifying areas of interest.

[0082] S330: Identify a region of interest in the target video stream according to the target boundary recognition model.

[0083] By inputting the target video stream into the target boundary recognition model, the region of interest can be obtained.

[0084] In an optional embodiment, before identifying the region of interest in the target video stream, the method further includes: preprocessing the video frames in the target video stream, and storing the preprocessed video frames as enhanced sample data for improving recognition accuracy.

[0085] Preprocessing can be normalizing the video frames to facilitate the subsequent target boundary recognition model to identify the region of interest. Exemplarily, preprocessing can include image denoising, image enhancement, image correction, size adjustment, and color space conversion, etc., which are not specifically limited in this application.

[0086] After the video frames are preprocessed, the original image information is synchronously output and archived to form enhanced sample data, which is used for subsequent supplementary reinforcement learning of the target boundary recognition model to improve the accuracy of the recognition results of the target boundary recognition model.

[0087] Before identifying the area of ​​interest in the target video stream, it also includes: preprocessing the video frames in the target video stream, standardizing the image format, and storing the preprocessed video frames as enhanced sample data to improve the recognition accuracy, increase the training samples of the target boundary recognition model, and improve the recognition accuracy of the target boundary recognition model.

[0088] S340, identifying solid waste in the area of ​​interest, and determining inventory parameters of the identified solid waste.

[0089] The technical solution of this embodiment determines the target boundary recognition model from the candidate boundary recognition models according to the storage boundary type of the warehouse to be monitored, and determines the target boundary recognition model according to the storage boundary type, so that the target boundary recognition model can more accurately identify the corresponding region of interest; according to the target boundary recognition model, the region of interest in the target video stream is identified, the accuracy of the region of interest is improved, and the subsequent occupancy of computing resources is reduced.

[0090] Embodiment 4

[0091] Figure 4 The figure shows a schematic diagram of the structure of a solid waste monitoring device provided in the fourth embodiment of the present application. This embodiment can be applied to the situation of monitoring the inventory data of solid waste in a warehouse storing solid waste. The specific structure of the solid waste monitoring device is as follows:

[0092] The target video stream acquisition module 410 is used to acquire the target video stream of the warehouse to be monitored according to a preset analysis cycle;

[0093] An area of ​​interest identification module 420, for identifying an area of ​​interest in a target video stream;

[0094] The inventory parameter determination module 430 is used to identify solid waste in the area of ​​interest and determine inventory parameters of the identified solid waste.

[0095] The technical solution of this embodiment obtains the target video stream of the warehouse to be monitored according to the preset analysis cycle, and monitors the warehouse through the video, without the need for manual on-site verification, thereby reducing labor costs; identifies the area of ​​interest in the target video stream, reduces the monitoring area of ​​solid waste, improves the accuracy of subsequent monitoring, and reduces the requirements for computing resources; identifies the solid waste in the area of ​​interest, and determines the inventory parameters of the identified solid waste, and directly monitors the solid waste in the warehouse to be monitored through the target video stream, thereby reducing the interference of factors such as vehicles, improving the accuracy of inventory parameters, and does not need to analyze the behavior of the vehicle, nor does it need to coordinate other data, thereby reducing the requirements for computing resources. Therefore, through the technical solution of this application, the problem that the transportation vehicles are relatively free, which may cause data errors, and the need to coordinate data from multiple parties, which has a high requirement for computing resources, is solved, and the accuracy of the inventory parameters of solid waste is improved, thereby improving the accuracy of the in-and-out data obtained according to the inventory parameters, and reducing the requirements for computing resources.

[0096] Optionally, the inventory parameter determination module 430 includes:

[0097] A solid waste identification unit, used to identify solid waste in an area of ​​interest by using a trained solid waste packaging identification model;

[0098] The inventory parameter determination unit is used to determine the inventory parameters of the identified solid waste according to the area parameters of the warehouse to be monitored and the identified solid waste, and the inventory parameters include at least one of the quantity and the floor area ratio.

[0099] Optionally, a solid waste packaging recognition model is trained by:

[0100] Label the packaging images of various types of solid waste to obtain a sample data set;

[0101] Build an initial solid waste packaging recognition model;

[0102] The initial solid waste packaging recognition model is trained according to the sample data set to obtain a trained solid waste packaging recognition model.

[0103] Optionally, the region of interest identification module 420 includes:

[0104] A target boundary recognition model determination unit, used to determine a target boundary recognition model from candidate boundary recognition models according to the storage boundary type of the warehouse to be monitored;

[0105] The region of interest recognition unit is used to recognize the region of interest in the target video stream according to the target boundary recognition model.

[0106] Optionally, the candidate boundary recognition model includes:

[0107] Overall warehouse recognition model, physical partition warehouse recognition model, boundary line recognition model and borderless warehouse recognition model.

[0108] Optionally, the solid waste monitoring device further comprises:

[0109] The enhanced sample data generation module is used to preprocess the video frames in the target video stream and store the preprocessed video frames as enhanced sample data to improve the recognition accuracy.

[0110] The solid waste monitoring device provided in the embodiments of the present application can execute the solid waste monitoring method provided in any embodiment of the present application, and has the corresponding functional modules and beneficial effects for executing the solid waste monitoring method.

[0111] Embodiment 5

[0112] Figure 5 A schematic diagram of the structure of an electronic device provided in Embodiment 5 of the present application is shown in FIG. Figure 5 As shown, the electronic device includes a processor 510, a memory 520, an input device 530 and an output device 540; the number of the processor 510 in the electronic device can be one or more. Figure 5 A processor 510 is taken as an example; the processor 510, the memory 520, the input device 530 and the output device 540 in the electronic device can be connected via a bus or other means. Figure 5 The example of connecting through bus is taken in the following.

[0113] The memory 520, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the solid waste monitoring method in the embodiment of the present application (for example, the target video stream acquisition module 410, the region of interest identification module 420 and the inventory parameter determination module 430). The processor 510 executes various functional applications and data processing of the electronic device by running the software programs, instructions and modules stored in the memory 520, that is, realizing the above-mentioned solid waste monitoring method.

[0114] The memory 520 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 520 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 520 may further include a memory remotely arranged relative to the processor 510, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0115] The input device 530 may be used to receive input character information and generate key signal input related to user settings and function control of the electronic device. The output device 540 may include a display device such as a display screen.

[0116] According to an embodiment of the present invention, the present invention also provides an electronic device, a readable storage medium and a computer program product.

[0117] Embodiment 6

[0118] Embodiment 6 of the present application also provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute a solid waste monitoring method, the method comprising: acquiring a target video stream of a warehouse to be monitored according to a preset analysis cycle; identifying an area of ​​interest in the target video stream; identifying solid waste in the area of ​​interest, and determining inventory parameters of the identified solid waste.

[0119] Of course, the storage medium containing computer executable instructions provided in the embodiment of the present application, whose computer executable instructions are not limited to the method operations described above, can also execute related operations in the solid waste monitoring method provided in any embodiment of the present application.

[0120] Through the above description of the implementation method, the technicians in the relevant field can clearly understand that the present application can be implemented with the help of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for an electronic device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0121] It is worth noting that in the above-mentioned embodiment of the solid waste monitoring device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application.

[0122] Note that the above are only preferred embodiments of the present application and the technical principles used. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the appended claims.

Claims

1. A solid waste monitoring method, characterized in that: include: According to the preset analysis cycle, obtain the target video stream of the warehouse to be monitored; Identifying a region of interest in the target video stream; Solid waste in the area of ​​interest is identified, and inventory parameters for the identified solid waste are determined.

2. The method according to claim 1, characterized in that The identifying solid waste in the area of ​​interest and determining inventory parameters of the identified solid waste includes: Identify the solid waste in the region of interest using a trained solid waste packaging recognition model; According to the area parameters of the warehouse to be monitored and the identified solid waste, the inventory parameters of the identified solid waste are determined, and the inventory parameters include at least one of quantity and floor space ratio.

3. The method according to claim 1, characterized in that The solid waste packaging recognition model is trained in the following way: Label the packaging images of various types of solid waste to obtain a sample data set; Build an initial solid waste packaging recognition model; The initial solid waste packaging recognition model is trained according to the sample data set to obtain a trained solid waste packaging recognition model.

4. The method according to claim 1, characterized in that: The identifying the region of interest in the target video stream comprises: Determining a target boundary recognition model from candidate boundary recognition models according to the storage boundary type of the warehouse to be monitored; According to the target boundary recognition model, a region of interest in the target video stream is identified.

5. The method according to claim 4, characterized in that The candidate boundary recognition model includes: Overall warehouse recognition model, physical partition warehouse recognition model, boundary line recognition model and borderless warehouse recognition model.

6. The method according to claim 1, characterized in that Before identifying the region of interest in the target video stream, the method further includes: The video frames in the target video stream are preprocessed, and the preprocessed video frames are stored as enhanced sample data for improving recognition accuracy.

7. A solid waste monitoring device, characterized in that: include: A target video stream acquisition module is used to acquire the target video stream of the warehouse to be monitored according to a preset analysis cycle; An area of ​​interest identification module, used to identify an area of ​​interest in the target video stream; The inventory parameter determination module is used to identify the solid waste in the area of ​​interest and determine the inventory parameters of the identified solid waste.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the solid waste monitoring method as described in any one of claims 1-6 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the solid waste monitoring method as described in any one of claims 1-6 is implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the solid waste monitoring method according to any one of claims 1 to 6.

Images