Laser three-dimensional real-time modeling and inventory management method and system for bulk materials in greenhouse

Abstract

The invention discloses a laser three-dimensional real-time modeling and inventory management method and system for bulk materials in a greenhouse. The system comprises a laser scanning unit; a weighing unit; a point cloud processing and modeling unit; an inventory management module; a data storage unit; a digital twin module; and the man-machine interaction unit is applied to warehousing, in-warehouse monitoring and checking, ex-warehouse and verification and alarm processes. According to the invention, through deep fusion of laser three-dimensional scanning and weighing sensing, not only is the high-precision material volume obtained, but also the real density is dynamically calculated, the industrial problem of inaccurate weight estimation caused by material density change is fundamentally solved, and the accuracy of inventory data is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of material control technology, and in particular to a laser 3D real-time modeling and inventory management method and system for bulk materials in greenhouses. Background Technology

[0002] Inventory management of bulk materials in greenhouses, such as grains, fertilizers, and feed, is increasingly becoming a key factor affecting operational efficiency and cost control. Traditional inventory management of bulk materials in greenhouses mainly relies on manual counting, static weighing, or experience-based estimation. However, traditional methods of volume measurement are inaccurate because the irregular stacking shape of bulk materials makes it difficult to accurately calculate the material volume using traditional size estimation or experience-based judgment methods. This leads to large errors in inventory statistics. Manual counting is time-consuming and inefficient, and it cannot achieve real-time updates of inventory data, affecting the timeliness of inbound and outbound decisions and inventory warnings. Moreover, the density of bulk materials may change during storage due to environmental factors such as humidity and compaction. Traditional fixed-density estimation methods will cause errors in weight estimation. Furthermore, managers cannot intuitively grasp the material stacking shape and inventory dynamics, making it difficult to achieve accurate in-warehouse scheduling and optimized utilization. For example, patent CN103913116B uses two-dimensional laser scanning to measure the material stack outline, but this method is complex to operate, slow in measurement speed, and has a low degree of automation, failing to meet the requirements of large-scale, high-frequency real-time inventory counting. Moreover, most current systems only use laser scanning to calculate volume, while weight still needs to be entered separately or estimated based on fixed theoretical density. This results in a large discrepancy between the weight measurement results and the actual weight, especially when the material's moisture content and compaction change, leading to significant errors in inventory management. Summary of the Invention

[0003] The present invention aims to address the shortcomings of the prior art by providing a laser three-dimensional real-time modeling and inventory management method and system for bulk materials in greenhouses.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A laser 3D real-time modeling and inventory management system for bulk materials in greenhouses includes:

[0006] Laser scanning unit: A three-dimensional laser scanner fixedly installed on the top of the greenhouse, used to scan bulk materials in the greenhouse in real time and generate point cloud data that characterizes the surface morphology of the materials.

[0007] Weighing unit: A weighing sensor installed at the bottom of the bulk cargo silo in the greenhouse or at a key weighing support structure, used to obtain the weight M of the bulk cargo.

[0008] Point cloud processing and modeling unit: Communicatively connected to the laser scanning unit and weighing unit, it receives point cloud data and weight data M, processes the point cloud data using a point cloud slicing algorithm, calculates the material volume V, and generates a 3D model of the material profile based on formulas. Calculate the real-time density of the material ;

[0009] Inventory management module: used for inventory management based on volume V, weight M, and density. It requires at least one parameter to perform inventory material receiving, issuing, inventory counting, and inventory level alerts;

[0010] Data storage unit: used to store point cloud data, 3D models, inventory records, and system parameters;

[0011] Digital twin module: Connected to the point cloud processing and modeling unit, it is used to build a virtual model consistent with the physical scene of the greenhouse based on point cloud data, and to drive the virtual model to update dynamically through real-time data;

[0012] Human-computer interaction unit: Connected to the digital twin module and inventory management module, it is used to receive user instructions and visually display virtual models, inventory data and early warning information to the user.

[0013] The laser scanning unit includes a single three-dimensional laser scanner, which is fixedly installed at a high point in the center or at one end of the greenhouse.

[0014] The laser scanning unit includes two or more 3D laser scanners, which are installed on different sides of the material pile. Through multi-view point cloud registration and fusion algorithms, a high-precision 3D model of the bulk material is generated.

[0015] The point cloud processing and modeling unit includes:

[0016] Point cloud fusion module: used to register point cloud datasets from different scanning angles and generate target point cloud datasets;

[0017] Dimensionality reduction module: Used to set the dimensionality reduction ratio according to the precision of the point cloud, perform covariance matrix dimensionality reduction on the target point cloud data, and output the dimensionality-reduced point cloud dataset;

[0018] Gaussian Modeling Module: Used to perform Gaussian process regression modeling on dimensionality-reduced point cloud datasets, outputting a surface fitting model of the material.

[0019] The data storage unit is also used to store historical point cloud data, 3D model versions, inventory change records, warning history and user operation logs, and supports time-series-based data query and analysis.

[0020] The digital twin module is also used to simulate the process of material warehousing, warehousing, and stacking, and dynamically updates the shape, volume, and position of the material in the virtual model based on real-time point cloud data, so as to realize the synchronous mapping between the physical scene and the virtual model.

[0021] A method for real-time laser 3D modeling and inventory management of bulk materials in greenhouses, utilizing a real-time laser 3D modeling and inventory management system for bulk materials in greenhouses, includes the following steps:

[0022] (1) Warehousing process:

[0023] Step S1: When new materials are transported into the shed for stacking, the three-dimensional laser scanner fixedly installed on the top of the shed is activated to scan the bulk materials in the shed in real time, collect point cloud data that characterizes the surface morphology of the materials, and simultaneously obtain the total weight M of the materials measured by the weighing sensor.

[0024] Step S2: Based on the initial point cloud data, a three-dimensional model of the material pile is generated and the total volume V of the material pile is calculated through three-dimensional reconstruction and point cloud slicing algorithms.

[0025] Step S3: According to the formula The average density of this batch of materials was calculated. And record the weight M, volume V, and density of this batch of materials. The 3D model is stored in the database as an entry record along with a timestamp;

[0026] (2) Inventory monitoring and counting process:

[0027] Step P1: At the set cycle or response time, the 3D laser scanner is activated to periodically scan all material piles in the shed to obtain the latest point cloud data.

[0028] Step P2: Based on the latest point cloud data, dynamically update the 3D model of each stockpile and recalculate its current volume. ;

[0029] Step P3: Retrieve the density recorded when the material was placed into the warehouse. According to the formula Estimate the weight of current inventory materials ;

[0030] Step P4: Estimate the current weight With volume As real-time inventory data updates, it is visualized on the human-computer interaction interface;

[0031] (3) Outbound process:

[0032] Step T1: When an outbound instruction is received, the outbound order information is parsed;

[0033] Step T2: Before and after the outbound operation, start the 3D laser scanner to scan the target material pile;

[0034] Step T3: By comparing the point cloud data and 3D model before and after the outbound shipment, the volume of material consumed in this outbound shipment is calculated. ;

[0035] Step T4: Based on the density of the stockpile Calculate the weight of the materials to be shipped out. And generate outbound records to update inventory;

[0036] (4) Verification and alarm process:

[0037] F1 step: Calculate the outbound weight based on the 3D model. The weight is compared with the expected weight on the outbound order. If the deviation exceeds a preset threshold, a verification alarm is triggered.

[0038] F2 step: Continuously monitor the changing trends of the volume and weight of each stockpile. When the inventory of any stockpile is lower than the preset safety threshold, an inventory warning will be automatically triggered.

[0039] The steps in step S2 described above, where the point cloud slicing algorithm processes point cloud data, are as follows:

[0040] E1 Step: Point Cloud Preprocessing

[0041] Before slicing the point cloud, denoising and coordinate correction preprocessing are performed to ensure data quality. Denoising: Remove discrete noise points caused by dust, sensor errors or flying insects. Coordinate correction: Ensure that the Z-axis of the point cloud is strictly parallel to the direction of gravity.

[0042] E2 Step: Determine the slice area and thickness:

[0043] and Find the minimum value of the point cloud of the entire stockpile on the Z-axis. (Usually the ground) and maximum value (the highest point of the pile), n;

[0044] Determine the slice thickness dz: Based on the density of the point cloud and the required computational accuracy, set a reasonable slice thickness. ;

[0045] E3 Step: Horizontal Layer Slicing

[0046] from Start to End, at height intervals Create a series of equally spaced horizontal cutting planes;

[0047] For each height : Where i = 1, 2, 3, ..., N... ;

[0048] Step E4: Process single-layer slices and calculate cross-sectional area: For the i-th slice, calculate the projected area of ​​that layer.

[0049] Extracting point cloud slices: Select all heights within the slice. Points within a thin layer;

[0050] Planar projection: Projecting all three-dimensional points (x, y, z) within the thin layer vertically onto a horizontal plane. Above, we obtain a set of two-dimensional points (x, y);

[0051] Calculate the projected area For a polygon with n vertices (x1, y1), (x2, y2), ..., (xn, yn), its area is: ;

[0052] Step E5: Integrate and sum to calculate the total volume:

[0053] By adding up the volumes of all the thin sheets, we can obtain the total volume V of the stockpile: ;

[0054] in This refers to the height of the highest point of the material pile.

[0055] In the aforementioned inventory monitoring and counting process, the system is also configured to perform density tracking and correction, specifically including:

[0056] The latest point cloud data and weighing sensor data are acquired synchronously on a regular basis or when triggered by a specific event.

[0057] According to the latest weight obtained simultaneously With the latest volume Recalculate the current density of the material. ;

[0058] Will Density recorded upon entry into the warehouse The comparison is performed, and if the relative rate of change exceeds a set threshold, then... This will serve as the new effective density for the stockpile, used for subsequent weight estimation.

[0059] Following step S3 of the warehousing process, a material identification assignment step is also included: assigning a unique identification code to each batch of incoming materials and associating the identification code with the corresponding weight. ,volume ,density It also includes 3D model association storage, which facilitates subsequent tracking and management.

[0060] Following step T4 of the outbound process, an outbound verification step is also included: displaying the changes in the 3D model of the outbound materials and the outbound weight through a human-computer interaction unit. The system requires users to confirm the outbound operation, and then updates the inventory record after confirmation.

[0061] The beneficial effects of this invention are as follows: By deeply integrating laser 3D scanning and weighing sensing, this invention not only obtains high-precision material volume but also dynamically calculates the true density, fundamentally solving the industry problem of inaccurate weight estimation caused by changes in material density. This significantly improves the accuracy of inventory data. Based on digital twin technology, a dynamic virtual model synchronized with the physical objects is created, enabling managers to remotely and intuitively grasp the precise shape and quantity of materials in the warehouse. It can automatically and quickly complete warehousing, inventory, and outbound operations, avoiding cumbersome and inefficient manual measurement and achieving 24 / 7 real-time monitoring and data updates. Attached Figure Description

[0062] Figure 1 This is a framework diagram of the laser 3D real-time modeling and inventory management system for bulk materials in greenhouses according to the present invention.

[0063] Figure 2 This is a flowchart of the laser three-dimensional real-time modeling and inventory management method for bulk materials in greenhouses according to the present invention.

[0064] The following will describe in detail, with reference to the accompanying drawings, embodiments of the invention. Detailed Implementation

[0065] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0066] A laser 3D real-time modeling and inventory management system for bulk materials in greenhouses includes:

[0067] Laser scanning unit: A three-dimensional laser scanner fixedly installed on the top of the greenhouse, used to scan bulk materials in the greenhouse in real time and generate point cloud data that characterizes the surface morphology of the materials.

[0068] Weighing unit: A weighing sensor installed at the bottom of the bulk cargo silo in the greenhouse or at a key weighing support structure, used to obtain the weight M of the bulk cargo.

[0069] Point cloud processing and modeling unit: Communicatively connected to the laser scanning unit and weighing unit, it receives point cloud data and weight data M, processes the point cloud data using a point cloud slicing algorithm, calculates the material volume V, and generates a 3D model of the material profile based on formulas. Calculate the real-time density of the material ;

[0070] Inventory management module: used for inventory management based on volume V, weight M, and density. It requires at least one parameter to perform inventory material receiving, issuing, inventory counting, and inventory level alerts;

[0071] Data storage unit: used to store point cloud data, 3D models, inventory records, and system parameters;

[0072] Digital twin module: Connected to the point cloud processing and modeling unit, it is used to build a virtual model consistent with the physical scene of the greenhouse based on point cloud data, and to drive the virtual model to update dynamically through real-time data;

[0073] Human-computer interaction unit: Connected to the digital twin module and inventory management module, it is used to receive user instructions and visually display virtual models, inventory data and early warning information to the user.

[0074] The laser scanning unit includes a single three-dimensional laser scanner, which is fixedly installed at a high point in the center or at one end of the greenhouse.

[0075] The laser scanning unit includes two or more 3D laser scanners, which are installed on different sides of the material pile. Through multi-view point cloud registration and fusion algorithms, a high-precision 3D model of the bulk material is generated.

[0076] The point cloud processing and modeling unit includes:

[0077] Point cloud fusion module: used to register point cloud datasets from different scanning angles and generate target point cloud datasets;

[0078] Dimensionality reduction module: Used to set the dimensionality reduction ratio according to the precision of the point cloud, perform covariance matrix dimensionality reduction on the target point cloud data, and output the dimensionality-reduced point cloud dataset;

[0079] Gaussian Modeling Module: Used to perform Gaussian process regression modeling on dimensionality-reduced point cloud datasets, outputting a surface fitting model of the material.

[0080] The data storage unit is also used to store historical point cloud data, 3D model versions, inventory change records, warning history and user operation logs, and supports time-series-based data query and analysis.

[0081] The digital twin module is also used to simulate the process of material warehousing, warehousing, and stacking, and dynamically updates the shape, volume, and position of the material in the virtual model based on real-time point cloud data, so as to realize the synchronous mapping between the physical scene and the virtual model.

[0082] A method for real-time laser 3D modeling and inventory management of bulk materials in greenhouses, utilizing a real-time laser 3D modeling and inventory management system for bulk materials in greenhouses, includes the following steps:

[0083] (1) Warehousing process:

[0084] Step S1: When new materials are transported into the shed for stacking, the three-dimensional laser scanner fixedly installed on the top of the shed is activated to scan the bulk materials in the shed in real time, collect point cloud data that characterizes the surface morphology of the materials, and simultaneously obtain the total weight M of the materials measured by the weighing sensor.

[0085] Step S2: Based on the initial point cloud data, a 3D model of the material pile is generated and the total volume V of the material pile is calculated through 3D reconstruction and point cloud slicing algorithms; the point cloud slicing algorithm in Step S2 processes the point cloud data as follows:

[0086] E1 Step: Point Cloud Preprocessing

[0087] Before slicing the point cloud, denoising and coordinate correction preprocessing are performed to ensure data quality. Denoising: Remove discrete noise points caused by dust, sensor errors or flying insects. Coordinate correction: Ensure that the Z-axis of the point cloud is strictly parallel to the direction of gravity.

[0088] E2 Step: Determine the slice area and thickness:

[0089] and Find the minimum value of the point cloud of the entire stockpile on the Z-axis. (Usually the ground) and maximum value (the highest point of the pile), n;

[0090] Determine the slice thickness dz: Based on the density of the point cloud and the required computational accuracy, set a reasonable slice thickness. ;

[0091] E3 Step: Horizontal Layer Slicing

[0092] from Start to End, at height intervals Create a series of equally spaced horizontal cutting planes;

[0093] For each height : Where i = 1, 2, 3, ..., N... ;

[0094] Step E4: Process single-layer slices and calculate cross-sectional area: For the i-th slice, calculate the projected area of ​​that layer.

[0095] Extracting point cloud slices: Select all heights within the slice. Points within a thin layer;

[0096] Planar projection: Projecting all three-dimensional points (x, y, z) within the thin layer vertically onto a horizontal plane. Above, we obtain a set of two-dimensional points (x, y);

[0097] Calculate the projected area For a polygon with n vertices (x1, y1), (x2, y2), ..., (xn, yn), its area is: ;

[0098] Step E5: Integrate and sum to calculate the total volume:

[0099] By adding up the volumes of all the thin sheets, we can obtain the total volume V of the stockpile: ;in This refers to the height of the highest point of the material pile.

[0100] Step S3: According to the formula The average density of this batch of materials was calculated. And record the weight M, volume V, and density of this batch of materials. The 3D model is stored in the database as an entry record along with a timestamp;

[0101] Following step S3 of the inbound process, there is also a material identification assignment step: assigning a unique identification code to each batch of inbound materials and associating the identification code with the corresponding weight. ,volume ,density It also includes 3D model association storage, which facilitates subsequent tracking and management.

[0102] (2) Inventory monitoring and counting process:

[0103] Step P1: At the set cycle or response time, the 3D laser scanner is activated to periodically scan all material piles in the shed to obtain the latest point cloud data.

[0104] Step P2: Based on the latest point cloud data, dynamically update the 3D model of each stockpile and recalculate its current volume. ;

[0105] Step P3: Retrieve the density recorded when the material was placed into the warehouse. According to the formula Estimate the weight of current inventory materials ;

[0106] Step P4: Estimate the current weight With volume As real-time inventory data updates, it is visualized on the human-computer interaction interface;

[0107] In the aforementioned inventory monitoring and counting process, the system is also configured to perform density tracking and correction, specifically including:

[0108] The latest point cloud data and weighing sensor data are acquired synchronously on a regular basis or when triggered by a specific event.

[0109] According to the latest weight obtained simultaneously With the latest volume Recalculate the current density of the material. ;

[0110] Will Density recorded upon entry into the warehouse The comparison is performed, and if the relative rate of change exceeds a set threshold, then... This will serve as the new effective density for the stockpile, used for subsequent weight estimation.

[0111] (3) Outbound process:

[0112] Step T1: When an outbound instruction is received, the outbound order information is parsed;

[0113] Step T2: Before and after the outbound operation, start the 3D laser scanner to scan the target material pile;

[0114] Step T3: By comparing the point cloud data and 3D model before and after the outbound shipment, the volume of material consumed in this outbound shipment is calculated. ;

[0115] Step T4: Based on the density of the stockpile Calculate the weight of the materials to be shipped out. And generate outbound records to update inventory;

[0116] Following step T4, there is also an outbound verification step: the 3D model changes and outbound weight of the materials are displayed through a human-machine interface unit. The system requires users to confirm the outbound operation, and then updates the inventory record after confirmation.

[0117] (4) Verification and alarm process:

[0118] F1 step: Calculate the outbound weight based on the 3D model. The weight is compared with the expected weight on the outbound order. If the deviation exceeds the preset threshold (set to ±5%), a verification alarm is triggered.

[0119] F2 step: Continuously monitor the changing trends of the volume and weight of each stockpile. When the inventory of any stockpile is lower than the preset safety threshold, an inventory warning will be automatically triggered.

[0120] like Figure 1 As shown, a real-time laser 3D modeling and inventory management system for bulk materials in a greenhouse includes a data acquisition layer, a point cloud processing and modeling unit, a business logic layer, a digital twin module, a human-computer interaction unit, and a data storage unit. The data acquisition layer collects physical data of the bulk materials inside the greenhouse, providing raw input for subsequent processing. The acquisition layer includes a laser scanning unit and a weighing unit. The laser scanning unit can be a single 3D laser scanner installed at the center of the greenhouse roof, or at one end, or multiple 3D laser scanners installed on different sides of the material pile. The laser scanning unit performs real-time scanning of the material pile, generating points representing the surface morphology of the materials. Cloud data is simultaneously acquired through weighing sensors installed at the bottom of the silo or key support structures to obtain real-time total weight data M of the material pile. Point cloud data is output from the laser scanning unit, and weight data is output from the weighing unit; both are transmitted to the point cloud processing and modeling unit. This unit registers and fuses point cloud data from multiple 3D laser scanners to generate a unified target point cloud dataset. Dimensionality reduction using the covariance matrix reduces the amount of point cloud data, outputting a dimensionality-reduced point cloud dataset. Gaussian regression is then performed on the dimensionality-reduced point cloud to fit a continuous model of the material surface. Finally, a layered slicing algorithm is used to calculate the total volume V and density of the material pile. Based on the processing structure of point cloud processing and modeling units, the system manages the entire inventory lifecycle, including inbound, in-stock monitoring, outbound, and verification alarms. The inbound management submodule receives point cloud data from laser scanning and weight data M from weighing, calls the point cloud slicing algorithm to calculate volume V, and calculates density. An inbound record is generated and stored in the data storage unit. The record includes weight data M, volume V, and density. The system uses 3D models, timestamps, and unique identifiers to perform periodic or on-demand inventory checks in the inventory monitoring and counting module: It activates a 3D laser scanner to acquire the latest point cloud data, updates the 3D model of the stockpile, and calculates the current volume. Use the density at the time of storage. Estimate current weight Display real-time inventory data , Upon receiving an outbound command from the human-computer interaction unit, the outbound management module calculates the volume of material consumed in this outbound process by comparing the point cloud data before and after the outbound process with the 3D model. Combined with the density of the stockpile Calculate the weight of the materials to be shipped out. It also generates outbound records and updates inventory; compares outbound weights. If the deviation from the expected weight on the outbound order exceeds a threshold, a verification alarm is triggered. The system continuously monitors the volume and weight trends of each stockpile; when the inventory of any stockpile falls below a preset safety threshold, an inventory warning is automatically triggered. The latest weight is simultaneously obtained. With the latest volume Recalculate the current density of the material. ;Will Density recorded upon entry into the warehouse The comparison is performed, and if the relative rate of change exceeds a set threshold, then... The new effective density of the material pile is used for subsequent weight estimation. The digital twin module constructs a 3D virtual model of the greenhouse material pile based on point cloud data, simulating the material warehousing, outbound, and stacking process. It receives real-time point cloud data and inventory changes, dynamically updates the virtual model, and achieves synchronous mapping between materials and the virtual model. The virtual model is then transmitted to the human-machine interaction module for display, and the virtual model version is stored in the data storage unit. The human-machine interaction unit transmits received user instructions, such as initiating inventory, executing outbound, and setting safety thresholds, to the inventory management module and data visualization module or the laser scanning unit and weighing unit. The human-machine interaction unit displays the digital twin virtual model of the material pile shape changing in real time, inventory data of volume, weight, and density, historical trends of inventory increase and decrease curves, verification alarms for outbound weight deviations, and inventory warnings for insufficient inventory.

[0121] Example 1

[0122] Taking a complete material management cycle as an example:

[0123] 1) Warehousing: After a batch of new coal is transported into the shed and the vehicle unloads and leaves, the manager clicks "Warehouse Registration" on the interface of the human-machine interface unit. The system automatically triggers the laser scanning unit and the weighing unit to obtain point cloud data and weight. =1050t, the point cloud processing and modeling unit completes the modeling within 30 seconds, and the volume V=850 is calculated using the point cloud slicing algorithm. Then the density ρ is obtained as 1.235t / The system generates and stores records;

[0124] 2) Inventory Monitoring and Counting: One week later, the system automatically conducts an inventory count at night, periodically scanning all material piles in the shed with a 3D laser scanner to obtain the latest point cloud data; dynamically updating the 3D model of each material pile and recalculating its current volume. =800 If the storage density ρ is used directly, and the estimated weight is 988t, then the system will synchronously read the weighing sensor data as follows: =1000t, recalculate the current density. =1.25t / The change rate from the initial density to the storage density is approximately 1.2%, which does not exceed the set threshold of 2%. Therefore, the system continues to use the original density. If the change rate exceeds the limit, it will automatically update to the new effective density.

[0125] 3) Outbound and Verification: Upon receiving an order for 50 tons of goods to be outbound, scan the volume V1=800 before outbound. After being shipped out, the scanned volume V2 = 760t, and the consumed volume =40 Calculate the outbound weight =1.235×40=49.4t, which is 0.6t off from the order of 50t. It is within the tolerance range, so the verification is passed. The system generates an outbound record, and the inventory weight is updated to 1000-49.4=950.6t. The human-computer interaction unit will display a 3D model comparison before and after the outbound shipment for management personnel to confirm.

[0126] In the description of the invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention.

[0127] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of the invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0128] In this invention, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0129] The invention has been described above with reference to the accompanying drawings. Obviously, the specific implementation of the invention is not limited to the above-described manner. Any improvements made using the inventive concept and technical solution, or direct application to other situations without modification, are all within the scope of protection of the invention.

Claims

1. A laser 3D real-time modeling and inventory management system for bulk materials in greenhouses, characterized in that, include Laser scanning unit: A three-dimensional laser scanner fixedly installed on the top of the greenhouse, used to scan bulk materials in the greenhouse in real time and generate point cloud data that characterizes the surface morphology of the materials. Weighing unit: A weighing sensor installed at the bottom of the bulk cargo silo in the greenhouse or at a key weighing support structure, used to obtain the weight M of the bulk cargo. Point cloud processing and modeling unit: Communicatively connected to the laser scanning unit and weighing unit, it receives point cloud data and weight data M, processes the point cloud data using a point cloud slicing algorithm, calculates the material volume V, and generates a 3D model of the material profile based on formulas. Calculate the real-time density of the material ; Inventory management module: used for inventory management based on volume V, weight M, and density. It requires at least one parameter to perform inventory material receiving, issuing, inventory counting, and inventory level alerts; Data storage unit: used to store point cloud data, 3D models, inventory records, and system parameters; Digital twin module: Connected to the point cloud processing and modeling unit, it is used to build a virtual model consistent with the physical scene of the greenhouse based on point cloud data, and to drive the virtual model to update dynamically through real-time data; Human-computer interaction unit: Connected to the digital twin module and inventory management module, it is used to receive user instructions and visually display virtual models, inventory data and early warning information to the user.

2. The laser 3D real-time modeling and inventory management system for bulk materials in greenhouses according to claim 1, characterized in that, The laser scanning unit includes a single three-dimensional laser scanner, which is fixedly installed at a high point in the center or at one end of the greenhouse.

3. The laser 3D real-time modeling and inventory management system for bulk materials in greenhouses according to claim 1, characterized in that, The laser scanning unit includes two or more 3D laser scanners, which are installed on different sides of the material pile. Through multi-view point cloud registration and fusion algorithms, a high-precision 3D model of the bulk material is generated.

4. The laser three-dimensional real-time modeling and inventory management system for bulk materials in greenhouses according to claim 3, characterized in that, The point cloud processing and modeling unit includes: Point cloud fusion module: used to register point cloud datasets from different scanning angles and generate target point cloud datasets; Dimensionality reduction module: Used to set the dimensionality reduction ratio according to the precision of the point cloud, perform covariance matrix dimensionality reduction on the target point cloud data, and output the dimensionality-reduced point cloud dataset; Gaussian Modeling Module: Used to perform Gaussian process regression modeling on dimensionality-reduced point cloud datasets, outputting a surface fitting model of the material.

5. A laser 3D real-time modeling and inventory management system for bulk materials in greenhouses according to claim 4, characterized in that, The data storage unit is also used to store historical point cloud data, 3D model versions, inventory change records, warning history and user operation logs, and supports time-series-based data query and analysis.

6. A laser 3D real-time modeling and inventory management system for bulk materials in greenhouses according to claim 4, characterized in that, The digital twin module is also used to simulate the process of material warehousing, warehousing, and stacking, and dynamically updates the shape, volume, and position of the material in the virtual model based on real-time point cloud data, so as to realize the synchronous mapping between the physical scene and the virtual model.

7. A method for real-time laser 3D modeling and inventory management of bulk materials in greenhouses, utilizing the real-time laser 3D modeling and inventory management system for bulk materials in greenhouses as described in claim 2, 3, or 6, characterized in that, Includes the following steps: (1) Warehousing process: Step S1: When new materials are transported into the shed for stacking, the three-dimensional laser scanner fixedly installed on the top of the shed is activated to scan the bulk materials in the shed in real time, collect point cloud data that characterizes the surface morphology of the materials, and simultaneously obtain the total weight M of the materials measured by the weighing sensor. Step S2: Based on the initial point cloud data, a three-dimensional model of the material pile is generated and the total volume V of the material pile is calculated through three-dimensional reconstruction and point cloud slicing algorithms. Step S3: According to the formula The average density of this batch of materials was calculated. And record the weight M, volume V, and density of this batch of materials. The 3D model is stored in the database as an entry record along with a timestamp; (2) Inventory monitoring and counting process: Step P1: At the set cycle or response time, the 3D laser scanner is activated to periodically scan all material piles in the shed to obtain the latest point cloud data. Step P2: Based on the latest point cloud data, dynamically update the 3D model of each stockpile and recalculate its current volume. ; Step P3: Retrieve the density recorded when the material was placed into the warehouse. According to the formula Estimate the weight of current inventory materials ; Step P4: Estimate the current weight With volume As real-time inventory data updates, it is visualized on the human-computer interaction interface; (3) Outbound process: Step T1: When an outbound instruction is received, the outbound order information is parsed; Step T2: Before and after the outbound operation, start the 3D laser scanner to scan the target material pile; Step T3: By comparing the point cloud data and 3D model before and after the outbound shipment, the volume of material consumed in this outbound shipment is calculated. ; Step T4: Based on the density of the stockpile Calculate the weight of the materials to be shipped out. And generate outbound records to update inventory; T5 Step: Outbound Verification: Display the changes in the 3D model of the outbound materials and their outbound weight through the human-computer interaction unit. The system requires users to confirm the outbound operation, and then updates the inventory record after confirmation. (4) Verification and alarm process: F1 step: Calculate the outbound weight based on the 3D model. The weight is compared with the expected weight on the outbound order. If the deviation exceeds a preset threshold, a verification alarm is triggered. F2 step: Continuously monitor the changing trends of the volume and weight of each stockpile. When the inventory of any stockpile is lower than the preset safety threshold, an inventory warning will be automatically triggered.

8. The method for real-time laser 3D modeling and inventory management of bulk materials in greenhouses according to claim 7, characterized in that, The steps in step S2 described above, where the point cloud slicing algorithm processes point cloud data, are as follows: E1 Step: Point Cloud Preprocessing Before slicing the point cloud, denoising and coordinate correction preprocessing are performed to ensure data quality. Noise reduction: Remove discrete noise caused by dust, sensor errors, or flying insects; coordinate correction: ensure that the Z-axis of the point cloud is strictly parallel to the direction of gravity; E2 Step: Determine the slice area and thickness: and Find the minimum value of the point cloud of the entire stockpile on the Z-axis. and maximum value , n; Determine the slice thickness dz: Based on the density of the point cloud and the required computational accuracy, set a reasonable slice thickness. ; E3 Step: Horizontal Layer Slicing from Start to End, at height intervals Create a series of equally spaced horizontal cutting planes; For each height : Where i = 1, 2, 3, ..., N... ; Step E4: Process single-layer slices and calculate cross-sectional area: For the i-th slice, calculate the projected area of ​​that layer. Extracting point cloud slices: Select all heights within the slice. Points within a thin layer; Planar projection: Projecting all three-dimensional points (x, y, z) within the thin layer vertically onto a horizontal plane. Above, we obtain a set of two-dimensional points (x, y); Calculate the projected area For a polygon with n vertices (x1, y1), (x2, y2), ..., (xn, yn), its area is: ; Step E5: Integrate and sum to calculate the total volume: By adding up the volumes of all the thin sheets, we can obtain the total volume V of the stockpile: ; in This refers to the height of the highest point of the material pile.

9. A laser three-dimensional real-time modeling and inventory management method for bulk materials in greenhouses according to claim 8, characterized in that, In the aforementioned inventory monitoring and counting process, the system is also configured to perform density tracking and correction, specifically including: The latest point cloud data and weighing sensor data are acquired synchronously on a regular basis or when triggered by a specific event. According to the latest weight obtained simultaneously With the latest volume Recalculate the current density of the material. ; Will Density recorded upon entry into the warehouse The comparison is performed, and if the relative rate of change exceeds a set threshold, then... This will serve as the new effective density for the stockpile, used for subsequent weight estimation.

10. A method for real-time laser 3D modeling and inventory management of bulk materials in greenhouses according to claim 9, characterized in that, Following step S3 of the warehousing process, a material identification assignment step is also included: assigning a unique identification code to each batch of incoming materials and associating the identification code with the corresponding weight. ,volume ,density It also includes 3D model association storage, which facilitates subsequent tracking and management.

Images