Goods label identification method based on intelligent logistics warehouse management

Abstract

The invention relates to the technical field of image recognition, and particularly discloses a cargo label recognition method based on intelligent logistics warehouse management, which comprises the following steps: S1, collecting an image in a warehouse by taking a vertex of a three-dimensional grid as a standard point and any standard point as a collection point, and determining a target shape in the image; s2, collecting a to-be-detected image, and determining a to-be-determined shape in the to-be-detected image; determining a standard point corresponding to the to-be-detected image, determining a to-be-compared shape, and determining a comparison shape based on the Euclidean distance between the center of the to-be-compared shape and the center of the to-be-determined shape; and S3, determining the similarity between the comparison shape and the to-be-determined shape, and judging whether the cargo label is an abnormal label or not according to the similarity. According to the invention, the abnormal cargo label is recognized in advance and prompted, so that the inventory management is prevented from being delayed.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a cargo label recognition method based on smart logistics warehouse management. Background Art

[0002] Smart logistics warehouse is a new type of warehousing that combines modern information technology and automation technology with the goal of improving warehouse management efficiency and optimizing logistics operations. It uses the Internet of Things (IoT), artificial intelligence (AI), etc. to realize automated, intelligent and digital management of warehousing, aiming to reduce costs and improve efficiency.

[0003] In smart logistics warehouse management, cargo labels are an important tool for efficient tracking, management and data interaction. Cargo labels carry the core information of the goods. By scanning the cargo labels, you can quickly and accurately obtain cargo information, such as cargo category, batch number and transportation history. This information is essential for inventory management, order fulfillment, quality traceability and transportation optimization. However, in actual situations, due to environmental, material or pasting process reasons, the pasted labels may curl or deform, affecting the normal scanning of the labels, resulting in the inability to obtain accurate cargo information and delaying inventory management. Summary of the invention

[0004] The purpose of the present invention is to provide a cargo label recognition method based on smart logistics warehouse management to solve the following technical problems: Due to environmental, material or pasting process reasons, the pasted label may curl or deform, affecting the normal scanning of the label, resulting in the inability to obtain accurate cargo information and delaying inventory management.

[0005] The purpose of the present invention can be achieved through the following technical solutions: A cargo label recognition method based on smart logistics warehouse management includes the following steps: S1: Divide the warehouse into three-dimensional grids to obtain a plurality of three-dimensional squares, wherein the length, width and height of the three-dimensional squares are all preset values, and take the vertices of the three-dimensional squares as standard points, place cargo labels at the standard points, take the standard point a as a collection point, collect images in the warehouse at the collection points, pre-process the images, determine the contour information of the cargo labels in the pre-processed images based on an edge detection algorithm, determine the shape of the cargo labels based on the contour information, and mark them as target shapes; S2: At the current time, an image in the warehouse is collected and marked as an image to be detected, the image to be detected is preprocessed, contour information of the cargo label in the image to be detected after the preprocessing is determined based on an edge detection algorithm, the shape of the cargo label in the image to be detected is determined based on the contour information, and it is marked as a shape to be determined; Determine the standard point A corresponding to the position where the image to be detected is collected, mark the target shape corresponding to the standard point A as the shape to be compared, determine the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined, and use the shape to be compared corresponding to the shortest Euclidean distance as the comparison shape; S3: superimposing the comparison shape and the shape to be determined, recording the superimposed area, rotating the shape to be determined and / or the comparison shape, and determining the maximum superimposed area S; When the superposition area is determined to be the maximum superposition area S, the positional relationship between the comparison shape and the shape to be determined is determined, a number of monitoring points are set at equal intervals on the edge of the shape to be determined, the projection distance from the monitoring point to the edge of the comparison shape is determined, and the mean value DJZ of the projection distance is calculated; Calculate the similarity P=η*S / (DJZ*D), where D represents the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined when the overlapping area is the maximum overlapping area S; A similarity threshold Pys is set. When the similarity P≤Pys, the cargo label corresponding to the undetermined shape is marked as an abnormal label, and the user is prompted.

[0006] As a further solution of the present invention: in the step S2, when the position where the image to be detected is collected is not on the standard point, the distance from the standard point to the position where the image to be detected is collected is determined, and the standard point corresponding to the shortest distance F is used as the standard point A.

[0007] As a further solution of the present invention: in the step S3, in the process of calculating the mean of the projection distance, when the difference between the projection distance d and the mean DJZ is greater than or equal to a preset difference threshold, the following steps are performed: When the projection distance d is less than DJZ, the projection distance d is removed, and the mean DJZ is calculated again; When the projection distance d≥DJZ, the shortest projection distance is removed and the mean DJZ is calculated again.

[0008] As a further solution of the present invention: in the step S3, when the projection distance from the monitoring point to the edge of the contrast shape is not unique, the shortest distance to be determined is used as the projection distance, and the distance to be determined is the projection distance from the monitoring point to the edge of the contrast shape.

[0009] As a further solution of the present invention: in the step S1, when dividing the warehouse into three-dimensional grids, when the length and / or width and / or height of the three-dimensional grids are not the preset values, the corresponding three-dimensional grids are removed.

[0010] As a further solution of the present invention: in the step S1, the edge detection algorithm includes the Canny algorithm, the Sobel algorithm and the Prewitt algorithm.

[0011] As a further solution of the present invention: in the step S1, the process of preprocessing the image includes denoising and enhancement.

[0012] As a further solution of the present invention: in the step S3, when the maximum overlapping area S≤Smin, the cargo label corresponding to the undetermined shape is marked as an abnormal label, and Smin represents a preset maximum overlapping area threshold.

[0013] Beneficial effects of the present invention: In this solution, by dividing the area in the warehouse according to a preset grid and placing a cargo label at each standard point, a reference library is established, and the image at each standard point is used as a reference for the subsequent detection process. By shooting and recording images at different positions, the shape changes of the cargo labels at different shooting angles and positions can be obtained, so as to clarify the expected shape of the cargo labels at each standard point, and provide an accurate comparison benchmark for subsequent real-time image detection. These standard templates are used as references to help the system identify whether there are any shape anomalies in the cargo labels in the actual collected images, so as to determine whether the cargo labels are abnormal, and solve the problem of label shape differences caused by different camera angles and shooting positions. This calibration process can compensate for this difference, making subsequent detection more robust, and because the corresponding label shape has been defined for each standard point during calibration, it is achieved that in the process of determining abnormal labels, taking one image can determine whether all cargo labels in the image are abnormal; by taking an image in real time, and then finding the corresponding cargo labels according to the positions of the cargo labels in the image. Standard point A, based on the label shape that has been calibrated by standard point A, these calibrated shapes are used as the shapes to be compared, and the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined in the real-time image (that is, the shape of the actual label in the image) is calculated to match the two, and determine which standard shape each shape to be determined should be compared with. At the same time, each label is directly compared with the corresponding standard point in the calibration image according to its spatial position in the image. The calibration process provides a reference template for each subsequent label shape, and ensures that each label can be compared by matching the center position (determining the shortest Euclidean distance), so that after one image acquisition, it is possible to simultaneously determine whether multiple labels are normal based on the calibrated shape and position data, without the need to take multiple images separately or process the labels one by one; finally, the similarity between the shape to be determined and the comparison shape is calculated, and whether the shape to be determined is consistent with the expected standard shape is determined by superposition, rotation, calculation of projection distance, etc. Finally, by calculating the similarity P, it is determined whether the label is abnormal; it is worth noting that the superposition area is the area of ​​the overlapping part after superposition, which will not be repeated here. The present invention avoids delays in inventory management by identifying abnormal cargo labels in advance and providing prompts. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below in conjunction with the accompanying drawings.

[0015] Figure 1 It is a flow chart of a cargo label identification method based on smart logistics warehouse management of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] See also Figure 1 As shown, the present invention is a cargo label identification method based on smart logistics warehouse management, comprising the following steps: S1: Divide the warehouse into three-dimensional grids to obtain a plurality of three-dimensional squares, wherein the length, width and height of the three-dimensional squares are all preset values, and take the vertices of the three-dimensional squares as standard points, place cargo labels at the standard points, take the standard point a as a collection point, collect images in the warehouse at the collection points, pre-process the images, determine the contour information of the cargo labels in the pre-processed images based on an edge detection algorithm, determine the shape of the cargo labels based on the contour information, and mark them as target shapes; S2: At the current time, an image in the warehouse is collected and marked as an image to be detected, the image to be detected is preprocessed, contour information of the cargo label in the image to be detected after the preprocessing is determined based on an edge detection algorithm, the shape of the cargo label in the image to be detected is determined based on the contour information, and it is marked as a shape to be determined; Determine the standard point A corresponding to the position where the image to be detected is collected, mark the target shape corresponding to the standard point A as the shape to be compared, determine the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined, and use the shape to be compared corresponding to the shortest Euclidean distance as the comparison shape; S3: superimposing the comparison shape and the shape to be determined, recording the superimposed area, rotating the shape to be determined and / or the comparison shape, and determining the maximum superimposed area S; When the superposition area is determined to be the maximum superposition area S, the positional relationship between the comparison shape and the shape to be determined is determined, a number of monitoring points are set at equal intervals on the edge of the shape to be determined, the projection distance from the monitoring point to the edge of the comparison shape is determined, and the mean value DJZ of the projection distance is calculated; Calculate the similarity P=η*S / (DJZ*D), where D represents the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined when the overlapping area is the maximum overlapping area S; A similarity threshold Pys is set. When the similarity P≤Pys, the cargo label corresponding to the undetermined shape is marked as an abnormal label, and the user is prompted.

[0018] It should be noted that by dividing the area according to the preset grid in the warehouse and placing the cargo label at each standard point, a reference library is established. The image at each standard point is used as a reference for the subsequent detection process. By shooting and recording images at different positions, the shape changes of the cargo labels at different shooting angles and positions can be obtained, so as to clarify the expected shape of the cargo labels at each standard point, providing an accurate comparison benchmark for subsequent real-time image detection. These standard templates are used as references to help the system identify whether there are any abnormalities in the shape of the cargo labels in the actual collected images, so as to determine whether the cargo labels are abnormal, and solve the problem of label shape differences caused by different camera angles and shooting positions. This calibration process can compensate for this difference, making subsequent detection more robust. In addition, since the corresponding label shape has been defined for each standard point during calibration, it is possible to determine whether all cargo labels in the image are abnormal by shooting one image in the process of determining abnormal labels; by shooting an image in real time, and then finding the corresponding standard point A according to the position of the cargo label in the image. Based on the label shapes that have been calibrated by the standard point A, these calibrated shapes are used as the shapes to be compared, and the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined in the real-time image (that is, the shape of the actual label in the image) is calculated to match the two, and determine which standard shape each shape to be determined should be compared with. At the same time, each label is directly compared with the corresponding standard point in the calibration image according to its spatial position in the image. The calibration process provides a reference template for each subsequent label shape, and ensures that each label can be compared by matching the center position (determining the shortest Euclidean distance), so that after one image acquisition, it is possible to simultaneously determine whether multiple labels are normal based on the calibrated shape and position data, without the need to take multiple images separately or process the labels one by one; finally, the similarity between the shape to be determined and the comparison shape is calculated, and whether the shape to be determined is consistent with the expected standard shape is determined by superposition, rotation, calculation of projection distance, etc. Finally, by calculating the similarity P, it is determined whether the label is abnormal; it is worth noting that the superposition area is the area of ​​the overlapping part after superposition, which will not be repeated here.

[0019] In another preferred embodiment of the present invention, in step S2, when the position where the image to be detected is collected is not on the standard point, the distance from the standard point to the position where the image to be detected is collected is determined, and the standard point corresponding to the shortest distance F is used as the standard point A.

[0020] It is worth noting that in actual operation, due to the limitations of the shooting angle or position of the camera, the acquisition position of the image to be detected may not be exactly on the standard point. In this case, by calculating the distance between the standard point and the acquisition position, the standard point closest to the acquisition position can be found as the standard point A, thereby avoiding the failure or inaccuracy of shape comparison caused by randomly selecting standard points.

[0021] In another preferred embodiment of the present invention, in the step S3, when the difference between the projection distance d and the mean DJZ is greater than or equal to a preset difference threshold, the following steps are performed: When the projection distance d is less than DJZ, the projection distance d is removed, and the mean DJZ is calculated again; When the projection distance d≥DJZ, the shortest projection distance is removed and the mean DJZ is calculated again.

[0022] In another preferred embodiment of the present invention, in the step S3, when the projection distance from the monitoring point to the edge of the contrast shape is not unique, the shortest distance to be determined is used as the projection distance, and the distance to be determined is the projection distance from the monitoring point to the edge of the contrast shape.

[0023] It should be noted that in actual situations, the projection distance may not be unique, because in some cases, there may be multiple possible calculation results for the projection of the monitoring point to the edge. For example, when a monitoring point is close to multiple edges of the contrasting shape, multiple projection distances will appear; the shortest projection distance often represents the closest match, indicating that the monitoring point is closer to the edge of the contrasting shape, thus reflecting the similarity in shape.

[0024] In another preferred embodiment of the present invention, in the step S1, when dividing the warehouse into three-dimensional grids, when the length and / or width and / or height of the three-dimensional grids are not the preset values, the corresponding three-dimensional grids are removed.

[0025] In another preferred embodiment of the present invention, in step S1, the edge detection algorithm includes Canny algorithm, Sobel algorithm and Prewitt algorithm.

[0026] It is understandable that the Canny algorithm (Canny Edge Detection) is a classic edge detection algorithm. Its goal is to find the edges in the image, that is, the area where the brightness or color of the image changes dramatically. The Canny algorithm is widely used in the field of image processing and computer vision, and performs well in detection accuracy, noise suppression, and edge continuity. I will not go into details about the Sobel algorithm and the Prewitt algorithm.

[0027] In another preferred embodiment of the present invention, in step S1, the process of preprocessing the image includes denoising and enhancement.

[0028] It is worth noting that denoising is an important step in image processing, and its purpose is to remove noise from the image to make the image clearer. Noise can come from equipment limitations during shooting, uneven lighting conditions, interference during image transmission, etc. The presence of noise will affect the effect of edge detection and other subsequent image processing steps, leading to false detection and misidentification; image enhancement aims to improve the contrast, brightness or other features in the image to make the useful information in the image more prominent. Before edge detection, image enhancement can improve the visibility of target objects (such as cargo labels) to ensure that the detection algorithm can more easily identify the contours and edges of objects.

[0029] In another preferred embodiment of the present invention, in step S3, when the maximum overlapping area S≤Smin, the cargo label corresponding to the undetermined shape is marked as an abnormal label, and Smin represents a preset maximum overlapping area threshold.

[0030] It is worth noting that the maximum overlapping area is the maximum overlap area obtained under all possible overlapping positions and angles. The positions of the shape to be determined and the comparison shape are adjusted by rotation, translation, etc. to find the best overlap and record the maximum overlapping area. When the overlapping area is less than or equal to Smin, it means that the shapes of the two are very different, and the label may have changed or is abnormal. Therefore, the cargo label corresponding to the shape to be determined is marked as an abnormal label.

[0031] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A cargo label recognition method based on smart logistics warehouse management, characterized in that: The following steps are involved: S1: Divide the warehouse into three-dimensional grids to obtain a plurality of three-dimensional squares, wherein the length, width and height of the three-dimensional squares are all preset values, and take the vertices of the three-dimensional squares as standard points, place cargo labels at the standard points, take the standard point a as a collection point, collect images in the warehouse at the collection points, pre-process the images, determine the contour information of the cargo labels in the pre-processed images based on an edge detection algorithm, determine the shape of the cargo labels based on the contour information, and mark them as target shapes; S2: At the current time, an image in the warehouse is collected and marked as an image to be detected, the image to be detected is preprocessed, contour information of the cargo label in the image to be detected after the preprocessing is determined based on an edge detection algorithm, the shape of the cargo label in the image to be detected is determined based on the contour information, and it is marked as a shape to be determined; Determine the standard point A corresponding to the position where the image to be detected is collected, mark the target shape corresponding to the standard point A as the shape to be compared, determine the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined, and use the shape to be compared corresponding to the shortest Euclidean distance as the comparison shape; S3: superimposing the comparison shape and the shape to be determined, recording the superimposed area, rotating the shape to be determined and / or the comparison shape, and determining the maximum superimposed area S; When the superposition area is determined to be the maximum superposition area S, the positional relationship between the comparison shape and the shape to be determined is determined, a number of monitoring points are set at equal intervals on the edge of the shape to be determined, the projection distance from the monitoring point to the edge of the comparison shape is determined, and the mean value DJZ of the projection distance is calculated; Calculate the similarity P=η*S / (DJZ*D), where D represents the Euclidean distance between the center of the shape to be compared and the center of the shape to be determined when the overlapping area is the maximum overlapping area S; A similarity threshold Pys is set. When the similarity P≤Pys, the cargo label corresponding to the undetermined shape is marked as an abnormal label, and the user is prompted.

2. According to claim 1, a cargo label identification method based on intelligent logistics warehouse management is characterized in that: In the step S2, when the position where the image to be detected is collected is not on the standard point, the distance from the standard point to the position where the image to be detected is collected is determined, and the standard point corresponding to the shortest distance F is used as the standard point A.

3. The cargo label identification method based on intelligent logistics warehouse management according to claim 1 is characterized in that: In the step S3, in the process of calculating the mean of the projection distances, when the difference between the projection distance d and the mean DJZ is greater than or equal to a preset difference threshold, the following steps are performed: When the projection distance d is less than DJZ, the projection distance d is removed, and the mean DJZ is calculated again; When the projection distance d≥DJZ, the shortest projection distance is removed and the mean DJZ is calculated again.

4. The cargo label identification method based on intelligent logistics warehouse management according to claim 1 is characterized in that: In the step S3, when the projection distance from the monitoring point to the edge of the contrast shape is not unique, the shortest distance to be determined is used as the projection distance, and the distance to be determined is the projection distance from the monitoring point to the edge of the contrast shape.

5. The cargo label identification method based on intelligent logistics warehouse management according to claim 1 is characterized in that: In the step S1, during the process of dividing the warehouse into three-dimensional grids, when the length and / or width and / or height of the three-dimensional grid are not the preset values, the corresponding three-dimensional grid is removed.

6. The method for cargo label identification based on intelligent logistics warehouse management according to claim 1 is characterized in that: In the step S1, the edge detection algorithm includes the Canny algorithm, the Sobel algorithm and the Prewitt algorithm.

7. The method for cargo label identification based on intelligent logistics warehouse management according to claim 1 is characterized in that: In the step S1, the process of preprocessing the image includes denoising and enhancement.

8. The cargo label identification method based on intelligent logistics warehouse management according to claim 1 is characterized in that: In the step S3, when the maximum overlapping area S≤Smin, the cargo label corresponding to the undetermined shape is marked as an abnormal label, and Smin represents a preset maximum overlapping area threshold.

Images