A computer vision-based rack inventory system and inventory method

Abstract

The application provides a computer vision-based vertical warehouse inventory system, comprising: a storage mechanism having a storage location; a stacker having a loading platform, the stacker being capable of transporting goods on the storage location to the loading platform; and an inventory device capable of recording the quantity of goods on the loading platform. The application discloses a computer vision-based vertical warehouse inventory system and inventory method, which has high modularity, fewer mechanical module sensors, simple installation, saves a large amount of space, and realizes efficient and high-precision inventory in a limited space in the vertical warehouse.

Description

Technical Field

[0001] This invention relates to the field of inventory counting, and in particular to a computer vision-based automated warehouse inventory counting system and method. Background Technology

[0002] With rapid economic growth, the demand for warehousing has also increased, making the development of intensive and automated warehousing an inevitable trend. Automated warehouses (AS / RS), with their large storage capacity and high space utilization, have become a key link in warehousing and logistics. However, AS / RS involve numerous interconnected devices and complex inbound and outbound processes, making errors and malfunctions inevitable during operation. If not handled promptly and properly, these can lead to lost or damaged goods, and discrepancies between system information and actual goods, resulting in significant economic losses for high-value industries such as pharmaceuticals, tobacco, and military products.

[0003] Currently, inventory counting technologies are mainly divided into three types: RFID inventory, weight inventory, and computer vision inventory. RFID wireless identification technology involves deploying RFID readers in designated areas within a logistics control system and attaching RFID tags to goods or pallets. Each time a pallet or goods passes by an RFID reader, the reader automatically scans the information on the tags and inputs the data into a data management information system for storage, analysis, and processing, thereby achieving the goal of controlling logistics. Weight inventory technology uses weighing devices to collect the weight information of materials. By comparing the weight difference before and after storage, it confirms the type, quantity, and other information of the materials. Computer vision inventory technology uses cameras to obtain point cloud data of the target, and uses computers to process the data to obtain information such as the size, volume, and placement of the goods, thus achieving inventory counting.

[0004] However, RFID inventory technology relies too heavily on the accuracy of tag information and the anti-interference ability of scanning equipment in complex electromagnetic environments, making it impossible to monitor the status of goods in real time. In practical applications, weight counting technology is severely affected by factors such as moisture in cartons and materials, which seriously impacts the accuracy of weight counting. Computer vision inventory methods require the use of sensors to scan goods from all angles. The scanning process has certain time and space requirements. In the aisles of automated warehouses, the space is small, limiting the deployment of sensors. Furthermore, the inventory counting process cannot interfere with the normal inbound and outbound processes, so it can only be carried out during business breaks, requiring high efficiency. Summary of the Invention

[0005] The purpose of this invention is to provide a computer vision-based automated warehouse inventory system to achieve fast and accurate inventory counting of goods.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a computer vision-based automated warehouse inventory system, comprising: a storage mechanism having storage locations; a stacker crane having a platform, the stacker crane being able to transport goods from the storage locations to the platform; and an inventory counting device being able to record the quantity of goods on the platform.

[0007] Furthermore, the storage mechanism includes a vertical storage frame, which is perpendicular to the horizontal plane. One side of the vertical storage frame has multiple storage positions along its length, and the stacker crane is located on the other side of the vertical storage frame, which can move along the side of the vertical storage frame.

[0008] Furthermore, the inventory equipment includes: a profile support, which is disposed on the platform; and a camera clamp, which is disposed on one side of the profile support, and a side camera is provided on the camera clamp, with the lens of the side camera facing the profile support.

[0009] Furthermore, the inventory equipment also includes a lifting mechanism, a clamping module, and a top-mounted camera; the lifting mechanism is connected to the profile support, the clamping module is connected to the lifting mechanism, and the top-mounted camera is connected to an extension mechanism. The lifting mechanism can drive the extension module to move vertically, and the lens of the top-mounted camera is located above the profile support and facing the profile support.

[0010] Furthermore, the inventory equipment also includes an electrical control box, which is connected to the lifting mechanism, the side camera, and the top camera.

[0011] Furthermore, there are multiple vertical warehouse supports, and each vertical warehouse support is equipped with a stacker crane.

[0012] Furthermore, it also includes a fisheye camera, which is mounted on the stacker crane and connected to the electrical control box. The fisheye camera is capable of capturing images of the storage location.

[0013] Furthermore, it also includes an industrial control computer, a wireless access point, a wireless repeater, and a server; the industrial control computer is connected to the electrical control box, the industrial control computer is connected to the wireless repeater through the wireless access point, and the wireless repeater is connected to the server.

[0014] On the other hand, a computer vision-based automated warehouse inventory method is provided, including the following steps: S1: Set the warehouse location to be inventoried; S2: Control the stacker crane to move to a warehouse location to be inventoried, and move the goods in the warehouse location to the platform of the stacker crane; S3: Take pictures of the goods on the platform from the side and top; Step S3 includes: S31: Set the target height; S32: Raise the camera above the goods to the target position; S33: Take pictures of the side and top of the goods through the camera on the side of the goods and the camera above the goods; S4: Obtain the quantity information, stack shape information, layer information and product specification information of the goods on the platform based on the captured pictures; S5: Repeat steps S2-S4 until all the goods in the warehouse locations to be inventoried have been photographed; S6: Determine whether the quantity information, stack shape information, layer information and product specification information of the goods are the same as the warehousing information.

[0015] Analysis shows that the present invention discloses a computer vision-based automated warehouse inventory system and method, which is highly modular, has fewer mechanical modules and sensors, is easy to install, saves a lot of space, and achieves efficient and high-precision inventory in the limited space of the automated warehouse. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. Wherein:

[0017] Figure 1 A schematic diagram of the structure of an inventory counting device according to an embodiment of the present invention.

[0018] Figure 2 A schematic diagram of the structure of an embodiment of the present invention.

[0019] Figure 3 A flowchart of an embodiment of the present invention.

[0020] Explanation of reference numerals in the attached diagram: 1. Profile bracket; 2. Counterweight plate; 3. Electrical control box; 4. Side camera clamp; 5. Industrial lens; 6. Side camera; 7. Servo motor; 8. Slide rail module; 9. Synchronous belt; 10. Reinforcing rib support structure; 11. First clamp module; 12. Connecting rod; 13. Second clamp module; 14. Top camera clamp; 15. Top camera; 16. Automated warehouse; 17. Automated warehouse support; 18. Storage location; 19. Goods; 20. Industrial control computer; 21. Fisheye camera; 22. Platform; 23. Stacker crane; 24. Wireless access point; 25. Wireless repeater; 26. Router; 27. Server. Detailed Implementation

[0021] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. Various examples are provided by way of explanation and not by way of limitation. Indeed, those skilled in the art will recognize that modifications and variations can be made to the invention without departing from its scope or spirit. For example, a feature shown or described as part of one embodiment may be used in another embodiment to produce yet another embodiment. Therefore, it is desirable that the invention encompass such modifications and variations falling within the scope of the appended claims and their equivalents.

[0022] In the description of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," and "bottom," etc., 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 do not require the invention to be constructed and operated in a specific orientation; therefore, they should not be construed as limitations on the invention. The terms "connected," "linked," and "set up" used in this invention should be interpreted broadly. For example, they can refer to a fixed connection or a detachable connection; a direct connection or an indirect connection through intermediate components; a wired connection, a radio connection, or a wireless communication signal connection. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0023] The accompanying drawings illustrate one or more examples of the invention. The detailed description uses numerals and letters to refer to features in the drawings. Similar or analogous reference numerals in the drawings and description have been used to refer to similar or analogous parts of the invention. As used herein, the terms “first,” “second,” “third,” and “fourth,” etc., are used interchangeably to distinguish one component from another and are not intended to indicate the location or importance of a single component.

[0024] like Figure 1 As shown, according to an embodiment of the present invention, a computer vision-based automated warehouse inventory system is provided, comprising: a storage mechanism having storage locations; a stacker crane 23 having a platform 22, the stacker crane 23 being able to transport goods 19 from storage locations to the platform 22; and inventory equipment being able to record the quantity of goods 19 on the platform 22.

[0025] Preferably, the storage mechanism includes a vertical storage frame 17, which is perpendicular to the horizontal plane. Multiple storage positions are provided on one side of the vertical storage frame 17 along its length. A stacker crane 23 is located on the other side of the vertical storage frame 17 and can move along the side of the vertical storage frame 17.

[0026] Preferably, the inventory equipment includes: a profile support 1, which is mounted on a platform 22; a camera clamp, which is mounted on one side of the profile support 1, and a side camera 6 mounted on the camera clamp, with the lens of the side camera 6 facing the profile support 1. The inventory equipment also includes a lifting mechanism, a clamping module, and a top camera 15; the lifting mechanism is connected to the profile support 1, the clamping module is connected to the lifting mechanism, and the top camera 15 is connected to the clamping module. The lifting mechanism can drive the clamping module to move vertically, and the lens of the top camera 15 is located above the profile support 1 and facing the profile support 1. The inventory equipment also includes an electrical control box 3, which is connected to the lifting mechanism, the side camera 6, and the top camera 15. The specific configuration of the inventory equipment is as follows: the profile support 1 is fixed on a stacker crane 23 in the vertical storage aisle, and a counterweight plate 2 is fixed to the bottom of the profile support 1 with screws to prevent the profile support 1 from shaking significantly during movement. The electrical control box 3 contains electrical components such as a PLC and relays, and is fixed to the rear end of the profile support 1. It is used to control the lifting mechanism, the side camera 6, and the top camera 15. The side camera 6 is an industrial 2D camera that captures side images of the goods 19 for stack matching, trademark matching, and text recognition. Typically, two side cameras 6 are arranged on the side, arranged sequentially along the vertical direction. The number of side cameras 6 can be increased according to actual usage requirements. The side camera 6 is equipped with an industrial lens 5 to ensure that the field of view covers the entire platform 22. It is fixed to the profile support 1 by the side camera 6 clamp 4, which can be quickly disassembled and installed to adjust the height of the side camera 6. The slide rail module 8 is fixed to the side of the profile support 1 with screws and driven by a servo motor 7. It is equipped with a synchronous belt 9 and together they form the lifting mechanism, thus solving the problem of not being able to arrange cameras in a confined space. The top-mounted camera 15 is a TOF camera. To ensure that the field of view covers the entire stack of goods 19, the top-mounted camera 15 must maintain a certain distance from the top of the goods 19. The use of a lifting mechanism can effectively ensure that the top-mounted camera 15 can collect relevant data, save a lot of space, and avoid interference between the crossbeams between the vertical warehouse and the top of the vertical warehouse and the top-mounted camera 15, making it safer and more reasonable. The top-mounted camera 15 is fixed on the top-mounted camera 15 clamp 14. The extension mechanism includes a second clamp module 13, a first clamp module 11, and a connecting rod 12. The top-mounted camera 15 clamp 14 is fixed to the connecting rod 12 through the second clamp module 13. The connecting rod 12 is connected to the lifting mechanism through the first clamp module 11. The reinforcing rib support structure 10 is used to improve the structural stability of the first clamp module 11 and the lifting mechanism. The second clamping module 13 and the first clamping module 11 are designed to adjust the top camera 15 to be in the center of the platform 22, and the camera's field of view is parallel to the profile support 1, so as to obtain the required image and point cloud data of the top surface of the goods 19 and realize the inventory function.

[0027] Preferably, such as Figure 2As shown, there are multiple automated storage and retrieval system (AS / RS) supports 17, each equipped with a stacker crane 23. A fisheye camera 21 is also included, mounted on the stacker crane 23 and connected to the electrical control box 3. The fisheye camera 21 can capture images of the storage locations 18. The system also includes an industrial control computer 20, a wireless access point 24, a wireless repeater 25, and a server 27. The industrial control computer 20 is connected to the electrical control box 3, and connects to the wireless repeater 25 via the wireless access point 24. The wireless repeater 25 is connected to the server 27. The entire AS / RS 16 contains multiple aisles. Multiple storage locations 18 are arranged vertically along the AS / RS supports 17 on both sides of the aisles. When the warehouse is full, goods 19 are stacked on the storage locations. A stacker crane 23 is installed on the other side of the AS / RS supports 17. The stacker crane 23 can move vertically along the AS / RS supports 17 to any storage location 18, enabling inbound, outbound, and inventory operations. When the inventory task is to scan empty storage location 18 or conduct an outbound inventory check, the stacker crane 23 moves to the target storage location, uses the fisheye camera 21 to photograph the situation inside storage location 18, transmits the data to the industrial control computer 20, and then moves the stacker crane 23 to the next storage location 18 to continue the above actions until the inventory check is completed. When the inventory task is an outbound inventory check, the stacker crane 23 removes the goods 19 from the pallet from storage location 18 and places them on the stacker crane 23 platform 22. At the same time, the PLC in the electrical control box 3 controls the lifting mechanism to rise to the designated position. After the above actions are completed, the side camera 6 and the top camera 15 take pictures to collect data, and transmit the data to the industrial control computer 20. Then, the PLC in the electrical control box 3 controls the lifting mechanism to retract, and the stacker crane 23 moves to the next storage location 18 to continue the above actions until the inventory check is completed. The PLC, industrial control computer 20, and wireless access point 24 are all fixed on the stacker crane 23 and move together with the stacker crane 23, together forming an electromechanical module. Wireless repeaters 25 are placed at the entrance of each alleyway, and servers 27 and routers 26 are deployed in the automated warehouse operation area. Wireless access points 24, wireless repeaters 25, and routers 26 form a mesh network to jointly constitute a communication module. When there is a signal, the data on the industrial control computer 20 and the results of the inventory algorithm are sent to the server 27. The server 27 compares the data with the warehouse's inbound and outbound information to jointly constitute a data processing module.

[0028] This invention also discloses a computer vision-based automated inventory method, such as... Figure 3As shown, the process includes the following steps: S1: Set the storage location 18 to be inventoried; S2: Control the stacker crane 23 to move to a storage location 18 to be inventoried, and move the goods 19 in the storage location 18 to the platform 22 of the stacker crane 23; S3: Take pictures of the goods 19 on the platform 22 from the side and top; Step S3 includes: S31: Set the target height; S32: Raise the camera above the goods 19 to the target position; S33: Take pictures of the side and top of the goods 19 through the camera on the side and the camera above the goods 19; S4: Obtain the quantity information, stack shape information, layer information and product specification information of the goods 19 on the platform 22 based on the pictures; S5: Repeat steps S2-S4 until all the goods 19 in the storage locations 18 to be inventoried have been photographed; S6: Determine whether the quantity information, stack shape information, layer information and product specification information of the goods 19 are the same as the warehousing information. The inventory task is divided into three inventory strategies: empty storage location 18 scanning, outbound inventory, and non-outbound inventory. For empty storage location 18 scanning: the server 27 selects the empty storage location 18 scanning mode and sends a scanning command via the local area network. The stacker crane 23 moves to the target storage location 18, and the fisheye camera 21 with a large field of view captures the situation inside storage location 18. The image is transmitted to the industrial control computer 20 via network cable. The image processing algorithm determines whether there are pallets and goods 19 in storage location 18, and writes the storage location 18 number and whether it is empty into the database. After scanning all target storage locations 18, the stacker crane 23 returns to the aisle entrance. At this time, the wireless access point 24 connects, and the database information is uploaded to the server 27 via the local area network. The server 27 reads the inbound and outbound information of the automated storage and retrieval system (AS / RS), compares the two to determine the empty storage location 18 number, and the scanned empty storage locations 18 are used for subsequent rapid replenishment of the AS / RS. Outbound Inventory: Select the outbound inventory mode on server 27 and send the inventory command via LAN. Stacker crane 23 moves to the target storage location 18. Stacker crane 23 takes out the pallet and goods 19 and places them on stacker crane platform 22. PLC controls the lifting mechanism to rise. Side camera 6 and top camera 15 capture the top and side information of goods 19 and transmit the images and point cloud data to industrial control computer 20. The side data is used to calculate information such as stack type, number of layers, and product specifications. The top data is used to calculate information such as stack type and quantity. The algorithm calculation results are written into the database. After scanning all target storage locations 18, stacker crane 23 returns to the aisle entrance. At this time, the wireless access point 24 signal is connected, and the database information is uploaded to server 27 via LAN. Server 27 reads the inbound and outbound information of the automated warehouse and compares the two to complete the inventory task.Inventory without leaving the warehouse: Select the inventory without leaving the warehouse mode on server 27, send the inventory command through the local area network, move the stacker crane 23 to the target storage location 18, use fisheye camera 21 to take pictures of the situation inside storage location 18, after the data is transmitted to industrial control computer 20, use visual algorithms such as image correction and image segmentation to obtain the quantity of goods 19 in storage location 18, write the calculation results into industrial control computer 20, and finally complete the inventory on server 27 in the same way.

[0029] From the above description, it can be seen that the above embodiments of the present invention achieve the following technical effects: In laboratory tests, the present invention can complete 18 inventory counts in a single warehouse location within 1 minute, with an inventory count accuracy rate of 98%, strong working stability, and high inventory count accuracy. The inventory system of the present invention has a high degree of modularity, fewer mechanical module sensors, simple installation, and saves a lot of space, achieving efficient and high-precision inventory counts in the limited space of an automated warehouse. The entire mechanical module, including the industrial control computer 20, electrical control box 3, wireless access point 24, lifting mechanism, and camera, is fixed together on the profile bracket 1. The profile bracket 1 can be directly combined with the stacker crane 23 platform 22 for installation, making it detachable, reusable, and easy to maintain. Multiple aisles can share a single set of mechanical modules, which can minimize costs. Except for the wired connection between mechanical modules, all other communication methods of the entire inventory system are wireless. Each module is simple and clear, with stronger logic, reduced redundancy, and reduced equipment complexity. The method of this invention integrates multiple functions and modes, and can perform empty warehouse location scanning, outbound inventory, outbound inventory, and inventory by designated lane or product specification. The main inventory information includes key information such as product specification, stack type, and quantity. It is fully functional and highly intelligent.

[0030] Compared with existing technologies, the vision-based automated warehouse inventory method of this invention integrates multiple modes, including empty warehouse location scanning, outbound inventory, non-outbound inventory, and inventory by product specification; it has multiple functions, including product specification detection, stack type recognition, and goods counting. This invention mainly utilizes computer vision technology to realize the inventory task, reconstructs the real scene through images and point cloud data, analyzes and calculates the target information, making the inventory more intelligent and stable. In this invention, the automated warehouse inventory device adopts a combination of a TOF camera, an industrial camera, and a fisheye camera 21. The TOF camera consists of an illumination unit, an optical lens, an imaging sensor, a control unit, and a computing unit. It continuously emits light pulses to the target, and then uses a sensor to receive the light returning from the object. The distance to the target object is obtained by detecting the flight time of the light pulses. The industrial 2D color camera has a high resolution of 12 million pixels, which can accurately capture the labels and text information on the goods 19. The fisheye camera 21 has a super-large field of view, which can acquire all the information of the goods 19 in a limited space. The inventory device for the neutral warehouse in this invention uses a lifting mechanism driven by a servo motor 7 to adjust the camera position. The entire lifting mechanism is controlled by a PLC. A proximity switch is installed at the end of the slide rail to prevent collisions. The effective travel of the slide rail is 1.3m, and a special reinforcing rib structure is designed to ensure sliding stability. Two clamping modules are designed to flexibly adjust the position and angle of the camera. The neutral warehouse inventory system in this invention uses wireless LAN communication to avoid the problem of difficult wiring in the neutral warehouse. A mesh network is achieved using a wireless access point 24 and a wireless repeater 25. When the signal is poor deep within the neutral warehouse and a communication connection cannot be established, the data is stored on the industrial control computer 20. After the stacker crane 23 returns to its origin, the signal is restored, and communication is normalized. The data on the industrial control computer 20 is then uploaded to the server 27. The mechanical modules of the neutral warehouse inventory system in this invention are easy to disassemble and can be reused in multiple aisles. The industrial control computer 20, electrical control box 3, wireless access point 24, lifting mechanism, and camera are all fixed on the profile bracket 1 and move together with the stacker crane 23. Operation is simple and stable.

[0031] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A computer vision-based racking inventory method, characterized by, The application relates to a warehouse inventory management system, which comprises the following parts: a warehouse inventory mechanism with storage locations; a stacker with a loading platform, which can transport goods on the storage locations to the loading platform; an inventory device which can record the quantity of goods on the loading platform; the warehouse inventory mechanism comprises a vertical storage support which is perpendicular to the horizontal plane, a plurality of storage locations are arranged on one side of the vertical storage support along the length direction, and the stacker is arranged on the other side of the vertical storage support and can move along the side of the vertical storage support; the inventory device comprises: a profile support arranged on the loading platform; a camera clamp arranged on one side of the profile support, wherein a side camera is arranged on the camera clamp, and the lens of the side camera faces the profile support; the inventory device further comprises a lifting mechanism, a clamping module and a top camera; the lifting mechanism is connected with the profile support, the clamping module is connected with the lifting mechanism, the top camera is connected with an extension mechanism, the lifting mechanism can drive the extension module to move in the vertical direction, and the lens of the top camera is located above the profile support and faces the profile support; the inventory device further comprises an electric control box which is connected with the lifting mechanism, the side camera and the top camera; the inventory device further comprises a fisheye camera arranged on the stacker, wherein the fisheye camera is connected with the electric control box, and the fisheye camera can shoot the storage locations; the application further comprises the following steps: S1: setting the storage locations which need to be inventoried; S2: controlling the stacker to move to one of the storage locations which need to be inventoried, and moving the goods in the storage location to the loading platform of the stacker; S3: shooting the goods on the loading platform from the side and the top; S4: obtaining the quantity information, the pile shape information, the layer information and the product specification information of the goods on the loading platform according to the shooting pictures; S5: repeating the steps S2-S4 until the goods in all the storage locations which need to be inventoried are shot; S6: judging whether the quantity information, the pile shape information, the layer information and the product specification information of the goods are the same as the storage information; the side camera adopts an industrial 2D camera, can collect the side images of the goods, and is used for pile type matching, trademark matching and character recognition; the top camera adopts a TOF camera, and the visual field range covers the whole pile of goods; the inventory task is divided into three inventory strategies, namely empty storage location scanning, warehouse-out inventory and non-warehouse-out inventory; in the empty storage location scanning, an empty storage location scanning mode is selected on a server, a scanning instruction is sent through a local area network, the stacker is moved to the side of a target storage location, a fisheye camera shoots the situation in the storage location, the image is transmitted to an industrial computer through a network cable, an image processing algorithm is used to judge whether there are pallets and goods in the storage location, a label of the storage location number and whether the storage location is empty is written into a database, when all the target storage locations are scanned, the stacker returns to the entrance of the lane, the database information is uploaded to the server through the local area network, the server reads the storage-in and storage-out information of the vertical storage, and the two are compared to determine the empty storage location number. Outbound inventory: Select the outbound inventory mode on the server, send the inventory instruction through the local area network, move the stacker to the target location, the stacker takes out the tray and goods and places them on the stacker platform, the PLC controls the lifting mechanism to rise, the side camera and the top camera shoot the top and side information of the goods, and the image and point cloud data are transmitted to the industrial computer, the side data is used to calculate the stack type, layer number and product specification information, the top data is used to calculate the stack type and quantity information, the algorithm calculation result is written into the database, when scanning all target locations, the stacker returns to the entrance of the lane, the database information is uploaded to the server through the local area network, the server reads the in-out information of the vertical warehouse, and the two are compared to complete the inventory task; Non-outbound inventory: Select the non-outbound inventory mode on the server, send the inventory instruction through the local area network, move the stacker to the target location, use the fisheye camera to shoot the situation in the location, and use the image correction and image segmentation visual algorithm to obtain the quantity of goods in the location after the data is transmitted to the industrial computer, and write the calculation result into the industrial computer to complete the inventory on the server.

2. The computer vision-based racking inventory method of claim 1, wherein, The number of the vertical warehouse supports is multiple, and each of the vertical warehouse supports is provided with a stacker.

3. A computer vision-based racking inventory method according to claim 2, wherein, It also includes an industrial computer, a wireless access point, a wireless repeater and a server. The industrial computer is connected with the electric control box, the industrial computer is connected with the wireless repeater through the wireless access point, and the wireless repeater is connected with the server.

4. The computer vision-based racking inventory method of claim 1, wherein, Step S3 includes: S31: Set the target height; S31: Raise the camera above the goods to the target position; S32: Shoot the side and top of the goods through the camera on the side of the goods and the camera above the goods.

Images