A warehouse management system based on real object ID

Abstract

The application belongs to the field of warehouse management, and discloses a warehouse management system based on physical ID, which comprises a conveyor belt, an infrared emitting device, an infrared receiving device, a shooting control device, a side camera and a top camera; the conveyor belt is used for transporting goods which need to be delivered out of or into the warehouse; the infrared emitting device is used for cooperating with the infrared receiving device to detect whether the goods are located in the shooting range of the top camera; the shooting control device is used for controlling the side camera to shoot the side image of the goods after receiving a first notification signal; the top camera is used for obtaining the upper surface image of the goods; the infrared receiving device is further used for sending a second notification signal to the shooting control device when infrared rays are received again; and the shooting control device is used for controlling the top camera to stop shooting after receiving the first notification signal. The application can improve the efficiency of delivering goods in and out of the warehouse.

Description

Technical Field

[0001] This invention relates to the field of warehouse management, and more particularly to a warehouse management system based on physical object IDs. Background Technology

[0002] In order to improve the efficiency of goods entering and leaving the warehouse, conveyor belts are usually used for transportation. The conveyor belt identifies the barcodes, QR codes and other labels affixed to the goods. Since the labels contain the physical ID of the goods, the inventory information can be automatically modified after the physical ID is identified, thus improving the efficiency of entering and leaving the warehouse.

[0003] In the process of label recognition for goods, existing technologies typically employ a triggered image capture method. This involves the camera taking a single shot when an object is detected being brought beneath it by a conveyor belt, capturing an image containing the label. However, this method does not account for the varying lengths of different goods, which can lead to variations in the label's placement on different objects. The single-shot method may result in missing images containing the label, leading to incorrect identification of the goods' physical ID. This necessitates manual processing for both outbound and inbound operations, increasing workload and reducing efficiency. Summary of the Invention

[0004] The purpose of this invention is to disclose a warehouse management system based on physical object ID, which solves the problem of how to increase the probability of obtaining images containing the labels affixed to the goods when photographing goods on a conveyor belt, thereby improving the efficiency of inbound and outbound operations.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] This invention provides a warehouse management system based on physical object ID, including a conveyor belt, an infrared transmitter, an infrared receiver, a shooting control device, a side camera, and a top camera;

[0007] Conveyor belts are used to transport goods that need to be shipped out or received. On the conveyor belt, the side of the goods with labels attached faces upwards.

[0008] An infrared emitting device is installed on one side of the conveyor belt to emit infrared rays horizontally in a direction perpendicular to the forward direction of the conveyor belt.

[0009] An infrared receiver is located on the other side of the conveyor belt to receive infrared rays emitted by the infrared transmitter and to send a first notification signal to the shooting control device when no infrared rays are received.

[0010] The top camera is positioned above the conveyor belt;

[0011] The side camera is set on one side of the conveyor belt and is used to take pictures in a direction perpendicular to the direction of the conveyor belt's movement.

[0012] The shooting control device is used to control the side camera to take a single shot after receiving a first notification signal to obtain a side image of the cargo; and to control the top camera to take vertical downward shots at an adaptive shooting interval based on the side image to obtain an image of the upper surface of the cargo.

[0013] The infrared receiver is also used to send a second notification signal to the shooting control device when infrared light is received again;

[0014] The shooting control device is used to control the top camera to stop shooting after receiving the first notification signal.

[0015] Preferably, it also includes a label recognition device;

[0016] The top camera is used to send the captured images of the upper surface to the tag recognition device;

[0017] The tag recognition device is used to identify the image on the upper surface and obtain the physical ID of the goods.

[0018] Preferably, it also includes an inventory management device;

[0019] The tag identification device is used to send the obtained physical item ID to the inventory management device;

[0020] The inventory management device is used to modify the inventory data stored in the database based on the physical item ID.

[0021] Preferably, it further includes an encoding device;

[0022] The coding device is used to generate physical IDs for goods that need to be put into storage.

[0023] Preferably, it further includes a generating device;

[0024] The generating device is used to generate tags based on the physical object ID.

[0025] Preferably, it also includes a printing device;

[0026] The printing device is used to print the labels generated by the generating device to obtain labels that can be pasted.

[0027] Preferably, the label includes a QR code or a barcode.

[0028] Preferably, the top camera is controlled to take vertically downward shots at an adaptive shooting interval based on the side image to obtain an image of the upper surface of the goods, including:

[0029] The side image is segmented to obtain a local image of the side of the cargo within the side image;

[0030] Calculate the height of the cargo based on local images;

[0031] Adaptive shooting interval based on altitude calculation;

[0032] The top camera is controlled to take a picture at regular intervals, resulting in multiple images of the upper surface of the goods.

[0033] Preferably, the side image is segmented to obtain a partial image of the side of the goods within the side image, including:

[0034] The side image is segmented using an image difference-based method to obtain a local image of the side of the cargo within the side image.

[0035] Preferably, calculating the height of the cargo based on a local image includes:

[0036] Calculate the height of the cargo using the following formula:

[0037]

[0038] H real The height of the cargo is represented by NL, and the total number of columns of pixels in the local image is represented by numpx. i The value represents the number of pixels in the i-th column of the local image, and prov represents the conversion ratio, which represents the height of one pixel in the side image in real space.

[0039] Beneficial effects:

[0040] This invention uses an infrared transceiver device installed on the top of the conveyor belt to detect whether the goods have entered or left the shooting range of the top camera. This allows for multiple shots of the goods while they are within the shooting range of the top camera, ensuring that all captured images include an image of the top surface of the label affixed to the goods. This effectively increases the probability of obtaining an image of the top surface containing the label, thereby improving the efficiency of inbound and outbound operations.

[0041] Furthermore, during the time interval between the first notification signal and the second notification signal received by the shooting control device, the top camera of the present invention does not use a fixed shooting interval to shoot the goods on the conveyor belt. Instead, it calculates an adaptive shooting interval using side images. This effectively increases the number of top surface images containing the label while reducing the total number of top surface images belonging to the same item. This reduces the workload of subsequent label recognition, which helps to reduce the energy consumption of the label recognition process and improve the efficiency of label recognition. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of a top view of a warehouse management system based on physical object ID according to the present invention.

[0044] Figure 2 This is a schematic diagram illustrating the connection relationship of a warehouse management system based on physical object ID according to the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] like Figure 1 As shown in one embodiment, the present invention provides a warehouse management system based on physical object ID, including a conveyor belt, an infrared transmitter, an infrared receiver, a shooting control device, a side camera, and a top camera;

[0047] Conveyor belts are used to transport goods that need to be shipped out or received. On the conveyor belt, the side of the goods with labels attached faces upwards.

[0048] An infrared emitting device is installed on one side of the conveyor belt to emit infrared rays horizontally in a direction perpendicular to the forward direction of the conveyor belt.

[0049] An infrared receiver is located on the other side of the conveyor belt to receive infrared rays emitted by the infrared transmitter and to send a first notification signal to the shooting control device when no infrared rays are received.

[0050] The top camera is positioned above the conveyor belt;

[0051] The side camera is set on one side of the conveyor belt and is used to take pictures in a direction perpendicular to the direction of the conveyor belt's movement.

[0052] The shooting control device is used to control the side camera to take a single shot after receiving a first notification signal to obtain a side image of the cargo; and to control the top camera to take vertical downward shots at an adaptive shooting interval based on the side image to obtain an image of the upper surface of the cargo.

[0053] The infrared receiver is also used to send a second notification signal to the shooting control device when infrared light is received again;

[0054] The shooting control device is used to control the top camera to stop shooting after receiving the first notification signal.

[0055] It's important to note that the top camera is not located on the conveyor belt, but rather above it, because... Figure 1 It is a top view.

[0056] The above solution uses an infrared transceiver device installed on the top of the conveyor belt to detect whether the goods have entered or left the shooting range of the top camera. This allows for multiple shots of the goods when they are within the shooting range of the top camera, ensuring that all captured images include an image of the top surface of the label affixed to the goods. This effectively increases the probability of obtaining an image of the top surface containing the label, thereby improving the efficiency of inbound and outbound operations.

[0057] Furthermore, during the time interval between the first notification signal and the second notification signal received by the shooting control device, the top camera of the present invention does not use a fixed shooting interval to shoot the goods on the conveyor belt. Instead, it calculates an adaptive shooting interval using side images. This effectively increases the number of top surface images containing the label while reducing the total number of top surface images belonging to the same item. This reduces the workload of subsequent label recognition, which helps to reduce the energy consumption of the label recognition process and improve the efficiency of label recognition.

[0058] Preferably, the infrared transmitter and receiver are positioned within the shooting range of the top camera, and the center of the side camera and the center of the top camera are in the same plane perpendicular to the surface of the conveyor belt.

[0059] Preferably, such as Figure 2 As shown, it also includes a label recognition device;

[0060] The top camera is used to send the captured images of the upper surface to the tag recognition device;

[0061] The tag recognition device is used to identify the image on the upper surface and obtain the physical ID of the goods.

[0062] When performing image recognition, the area where the label is located can be detected first, and then the label can be recognized to obtain the physical object ID contained in the label.

[0063] Preferably, the upper surface image is recognized to obtain the physical ID of the goods, including:

[0064] For the received k-th top surface image, obtain the time length between the reception time of the k-th top surface image and the reception time of the (k-1)-th top surface image;

[0065] If the time length is greater than the preset time threshold, it means that the kth top surface image belongs to another item.

[0066] For images of the upper surface of the same item, image recognition is performed sequentially according to the receiving time from earliest to latest. If the physical ID of the item is obtained during the recognition process, the remaining unrecognized upper surface images of the same item will no longer be recognized.

[0067] The above process can obtain images of the top surface of the same item, thereby improving the accuracy of the recognition results. Specifically, if the time interval between the reception of two received top surface images is greater than a preset time threshold, it indicates that the k-th image is an image of another item captured by the top camera.

[0068] Preferably, the preset duration threshold is the maximum value of the adaptive shooting interval.

[0069] Preferably, the process of recognizing the upper surface image includes:

[0070] The upper surface image is segmented into multiple region images containing the same number of pixels.

[0071] Calculate the recognition priority value of the region image;

[0072] Each region of the image is identified in descending order of recognition priority:

[0073] S1. Number each region image consecutively starting from 1 according to the recognition priority value. The larger the recognition priority value, the smaller the number.

[0074] S2, let z represent the number, initialize z = 1;

[0075] S3. For the z-th region image, obtain the image features of the region image, and determine whether there is a region belonging to the label in the z-th region image based on the image features. If yes, proceed to S5; otherwise, proceed to S4.

[0076] S4, increment the value of z by 1. If the value of z is less than or equal to the total number of region images, proceed to S3; if the value of z is greater than the total number of region images, it means that the upper surface image does not contain the label of the goods, and the calculation ends.

[0077] S5, determine whether the area belonging to the label is complete. If yes, proceed to S6; otherwise, proceed to S7.

[0078] S6, identify the area belonging to the label, obtain the physical ID represented by the label, and end the calculation;

[0079] S7, stitch all the region images adjacent to the region belonging to the label in the z-th region image with the z-th region image to obtain the stitched image;

[0080] S8. In the stitched image, obtain the complete area of ​​the label, identify the complete area, obtain the physical ID represented by the label, and end the calculation.

[0081] After segmentation, this invention does not identify regions according to their positional relationship within the upper surface image, as in existing technologies. Instead, it calculates recognition priority values ​​and identifies regions in descending order of priority, thus prioritizing regions with a higher probability of containing labels. This increases the probability of identifying the label region without needing to identify all regions in the upper surface image, effectively improving the recognition efficiency of the upper surface image.

[0082] The region where the label is located has line segment distribution characteristics and pixel distribution characteristics that are significantly different from other regions. By identifying these characteristics, it can be determined whether the region of the image contains a region belonging to the label.

[0083] Specifically, when recognizing a stitched image containing a complete region, the label region can be obtained first through image segmentation, and then the physical ID of the label can be obtained by recognizing the label region.

[0084] For example, when the label is a barcode, the physical ID represented by the barcode can be determined by obtaining the distance between the black stripes in the label area.

[0085] Preferably, the upper surface image is segmented into multiple region images containing the same number of pixels, including:

[0086] Let d represent the distance between the top surface of the cargo corresponding to the top surface image and the top camera;

[0087] Let N1 and N2 represent the length and width of the region image, respectively;

[0088] but

[0089] baseN1 represents the preset length; sd represents the distance between the top camera and the upper surface of the conveyor belt;

[0090]

[0091] baseN2 represents the preset length;

[0092] The upper surface image is then divided into multiple region images with a length of N1 and a width of N2.

[0093] In this invention, the size of the region image is not fixed, but adaptively changes with the distance *d* between the upper surface of the goods and the top camera. A larger *d* results in a smaller length and width of the region image; a smaller *d* results in a larger length and width. This allows the size of the region image to adaptively change with the proportion of the label in the upper surface image; a larger proportion results in a larger region image, and a smaller proportion results in a smaller region image. This helps avoid the situation where, at a large proportion, the number of region images is too large, leading to an excessive probability of the label's location being assigned to different region images, thus affecting label recognition efficiency. It also helps avoid the situation where, at a small proportion, the number of region images is too small, leading to reduced label recognition efficiency when using recognition priority values. In other words, the size of the region image in this invention is closely coordinated with the process of recognizing region images based on recognition priority values. This method of changing the size of the region image increases the probability of identifying regions belonging to the label in each region image earlier when recognizing each region image in descending order of recognition priority value, thereby improving the efficiency of identifying the label's location in the region image.

[0094] Preferably, the preset length is one-quarter of the length of the upper surface image.

[0095] Preferably, the preset width is one-quarter of the width of the upper surface image.

[0096] Preferably, the formula for calculating the priority value is:

[0097]

[0098] ideprival represents the recognition priority value, NU and NV represent the number of rows and columns of pixels in the region image, respectively, and pixelval. u,vdenoted as the gradient value of the pixel in row u and column v, avepixval represents the mean gradient value of the pixels in the region image, numline represents the number of straight line segments contained in the region image, and δ1 and δ2 represent the gradient coefficient and the straight line coefficient, respectively.

[0099] The recognition priority value of this invention considers two variables: the degree of gradient value fluctuation and the number of straight lines contained. The greater the fluctuation of the gradient value and the greater the number of straight lines contained, the higher the recognition priority value of the region image, indicating a higher probability that the image contains a region belonging to the label. Because goods are generally packaged when entering and leaving the warehouse, when the label is affixed to the surface of the packaging, the transition of the gradient value between the label area and other areas of the packaging in the obtained upper surface image may not be natural, i.e., there may be abrupt changes. This leads to increased gradient value fluctuation. Therefore, the greater the gradient value fluctuation, the greater the probability of containing the label. However, if only this situation is considered, when advertising slogans or patterns are printed in local areas of the packaging surface, the gradient value may also fluctuate significantly. In this case, it is impossible to correctly determine whether a region contains the label. Therefore, this invention incorporates the number of straight line segments, utilizing the characteristic that labels contain a large number of straight line segments, to further distinguish regions with similar gradient value fluctuations, thereby more accurately representing the probability that the region image contains a region belonging to the label.

[0100] Preferably, the gradient coefficient is 0.6 and the linearity coefficient is 0.4.

[0101] Preferably, determining whether the area belonging to the label is complete includes:

[0102] Let d represent the distance between the top surface of the cargo corresponding to the top surface image and the top camera;

[0103] Use proh d This represents the length in real space of one pixel in the captured image when the distance between the object being photographed and the top camera is d.

[0104] Get the smallest bounding rectangle of the region belonging to the label;

[0105] Get the number of pixels lenpx contained in the longer side of the smallest bounding rectangle;

[0106] The length lenA of the region belonging to the label in real space is:

[0107] lenA = proh d ×lenpx

[0108] If lenA is less than the actual length of the label lenB, it means that the area belonging to the label is incomplete.

[0109] The present invention determines whether a region is complete by comparing the length of the region with the actual length of the label. If lenA is less than lenB, it indicates that the region is not complete. In this case, adjacent region images need to be stitched together to obtain a stitched image containing the complete label.

[0110] Preferably, the z-th region image is stitched together with all region images adjacent to the region belonging to the label in the z-th region image to obtain a stitched image, including:

[0111] If the region belonging to the label in the z-th region image does not contain the pixel at the intersection of the two edges of the z-th region image, then there are three region images that are adjacent to the region belonging to the label in the z-th region image.

[0112] Otherwise, there is only one region image that is adjacent to the region belonging to the label in the z-th region image.

[0113] When the region belonging to the label contains pixels at the intersection of two edges of the z-th region image, it means that the region belonging to the label is located in a corner of the region image. In this case, the region of the label may be divided into 4 region images, so there are a total of 3 adjacent region images.

[0114] Preferably, it also includes an inventory management device;

[0115] The tag identification device is used to send the obtained physical item ID to the inventory management device;

[0116] The inventory management device is used to modify the inventory data stored in the database based on the physical item ID.

[0117] If the goods are currently in the outbound stage, the status of the goods in the inventory data will be changed to "outbound" based on the physical item ID; if the goods are currently in the inbound stage, the status of the goods in the inventory data will be changed to "inbound" based on the physical item ID.

[0118] Preferably, it further includes an encoding device;

[0119] The coding device is used to generate physical IDs for goods that need to be put into storage.

[0120] Specifically, the coding rules can be set according to actual needs. For example, the code can be based on the year, month, day, hour, minute, and second, plus a sequence code. For example, 202406012203001 represents the physical ID of the first item coded on June 1, 2024, at 22:03.

[0121] Preferably, it further includes a generating device;

[0122] The generating device is used to generate labels based on the physical object ID.

[0123] Specifically, it can generate QR codes or barcodes.

[0124] Preferably, it also includes a printing device;

[0125] The printing device is used to print the labels generated by the generating device to obtain labels that can be pasted.

[0126] Preferably, the label includes a QR code or a barcode.

[0127] Preferably, the top camera is controlled to take vertically downward shots at an adaptive shooting interval based on the side image to obtain an image of the upper surface of the goods, including:

[0128] The side image is segmented to obtain a local image of the side of the cargo.

[0129] Calculate the height of the cargo based on local images;

[0130] Adaptive shooting interval based on altitude calculation;

[0131] The top camera is controlled to take a picture at regular intervals, resulting in multiple images of the upper surface of the goods.

[0132] The top camera of this invention does not employ a fixed shooting interval. This is because, for goods with low heights, a fixed shooting interval would easily result in too many surface images that do not include the label, leading to excessive invalid images during the surface image recognition process and obviously reducing recognition efficiency. Conversely, for goods with high heights, a fixed shooting interval might result in two surface images taken at intervals not containing overlapping areas, potentially causing the label to be missed and resulting in label recognition failure. Therefore, the shooting method of this invention effectively improves the surface image recognition efficiency and significantly increases the success rate of label acquisition.

[0133] Preferably, the side image is segmented to obtain a partial image of the side of the goods within the side image, including:

[0134] The side image is segmented using an image difference-based method to obtain a local image of the side of the cargo within the side image.

[0135] Specifically, when segmenting the side image based on image difference, the side image is compared with the background image to obtain the set of pixels chgU in the side image whose gray values ​​have changed compared to the background image; the pixels in chgU form a local image of the side of the goods.

[0136] The background image is taken by a side camera when there are no goods on the conveyor belt. Therefore, by using image differencing, goods in the side image can be detected quickly.

[0137] Preferably, calculating the height of the cargo based on a local image includes:

[0138] Calculate the height of the cargo using the following formula:

[0139]

[0140] H real The height of the cargo is represented by NL, and the total number of columns of pixels in the local image is represented by numpx. i The value represents the number of pixels in the i-th column of the local image, and prov represents the conversion ratio, which represents the height of one pixel in the side image in real space.

[0141] By calculating the average number of pixels in each column, a more accurate height can be obtained, reducing measurement errors.

[0142] Preferably, the altitude-based adaptive shooting interval includes:

[0143]

[0144] shotT represents the adaptive shooting interval, and Tma represents the maximum preset shooting interval.

[0145] In this invention, the shooting interval is smaller when the upper surface of the goods is closer to the top camera, and larger when the upper surface of the goods is farther from the top camera, thus achieving adaptive variation of the shooting interval. This results in fewer upper surface images being obtained when the goods are farther from the top camera, improving recognition efficiency; and more upper surface images being obtained when the goods are closer to the top camera, increasing the probability of overlapping areas between adjacent upper surface images, thereby improving the success rate of tag recognition.

[0146] Specifically, the maximum shooting interval is related to the horizontal field of view of the top camera; the larger the horizontal field of view, the larger the maximum shooting interval.

[0147] Specifically, the maximum shooting interval is the maximum time length that satisfies the condition that there is at least 5% overlap in two images obtained by the top camera shooting vertically towards the conveyor belt at one shooting interval.

[0148] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A warehouse management system based on physical item IDs, characterized in that, Includes a conveyor belt, an infrared transmitter, an infrared receiver, a shooting control device, a side camera, and a top camera; Conveyor belts are used to transport goods that need to be shipped out or received. On the conveyor belt, the side of the goods with labels attached faces upwards. An infrared emitting device is installed on one side of the conveyor belt to emit infrared rays horizontally in a direction perpendicular to the forward direction of the conveyor belt. An infrared receiver is located on the other side of the conveyor belt to receive infrared rays emitted by the infrared transmitter and to send a first notification signal to the shooting control device when no infrared rays are received. The top camera is positioned above the conveyor belt; The side camera is set on one side of the conveyor belt and is used to take pictures in a direction perpendicular to the direction of the conveyor belt's movement. The shooting control device is used to control the side camera to take a single shot after receiving a first notification signal to obtain a side image of the cargo; and to control the top camera to take vertical downward shots at an adaptive shooting interval based on the side image to obtain an image of the upper surface of the cargo. The infrared receiver is also used to send a second notification signal to the shooting control device when infrared light is received again; The shooting control device is used to control the top camera to stop shooting after receiving the first notification signal; Based on the side images, the top camera is controlled to take vertical downward shots at an adaptive shooting interval to obtain images of the upper surface of the cargo, including: The side image is segmented to obtain a local image of the side of the cargo. Calculate the height of the cargo based on local images; Adaptive shooting interval based on altitude calculation; The top camera is controlled to take a picture at regular intervals to obtain multiple images of the upper surface of the goods; Calculating the height of cargo based on local images includes: Calculate the height of the cargo using the following formula: Indicates the height of the goods. This represents the total number of columns of pixels in a local image. This represents the number of pixels in the i-th column of a local image. This indicates the conversion ratio, which represents the height of one pixel in the side view image in real space.

2. The warehouse management system based on physical item ID according to claim 1, characterized in that, It also includes a label recognition device; The top camera is used to send the captured images of the upper surface to the tag recognition device; The tag recognition device is used to identify the image on the upper surface and obtain the physical ID of the goods.

3. A warehouse management system based on physical item ID according to claim 2, characterized in that, It also includes inventory management devices; The tag identification device is used to send the obtained physical item ID to the inventory management device; The inventory management device is used to modify the inventory data stored in the database based on the physical item ID.

4. A warehouse management system based on physical item ID according to claim 1, characterized in that, It also includes an encoding device; The coding device is used to generate physical IDs for goods that need to be put into storage.

5. A warehouse management system based on physical item ID according to claim 4, characterized in that, It also includes a generating device; The generating device is used to generate tags based on the physical object ID.

6. A warehouse management system based on physical item ID according to claim 5, characterized in that, It also includes a printing device; The printing device is used to print the labels generated by the generating device to obtain labels that can be pasted.

7. A warehouse management system based on physical item ID according to claim 1, characterized in that, Labels include QR codes or barcodes.

8. A warehouse management system based on physical item ID according to claim 1, characterized in that, The side image is segmented to obtain a local image of the side of the cargo within the side image, including: The side image is segmented using an image difference-based method to obtain a local image of the side of the cargo within the side image.

Images