A pallet automatic detection, counting and stacking device and warehouse supporting automation process

Abstract

The present invention proposes a pallet automatic detection, counting and fast stacking device and a warehouse supporting automation process, which includes a pallet placement box, a delivery port is arranged on the upper end surface of the pallet placement box, and visual statistical components and pallet deceleration buttons are respectively arranged on the opposite side surfaces of the pallet placement box, the visual statistical component includes a visual camera and a controller, the pallet deceleration button includes a deceleration ball, a spring and an adjustment seat, and also includes a reader and a counting display. An electronic identification code is installed on the pallet. The present application can ensure that the pallet is accurately aligned and stacked during the sliding process, detect the pallet status in real time, display the number of pallets, verify the delivery quantity, and integrate the pallet information in the pallet collection device into the WCS of the AGV, and can remotely adjust the AGV pallet picking strategy according to actual conditions, thereby greatly improving the efficiency and neatness of pallet stacking, improving delivery accuracy, and improving the safety of empty pallet storage and goods warehousing and shelving, ensuring the stable operation of the production line, and facilitating scientific decision-making and resource optimization allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics warehousing, and in particular to a pallet automatic detection, counting and stacking device and a warehousing matching automation process. Background Art

[0002] Pallets are important tools in the logistics, warehousing and supply chain industries. They have the advantages of being widely used in a wide range of industries, large in volume, low in cost, and can be used repeatedly. At the same time, they can also improve the efficiency of goods transportation, storage, loading and unloading, and reduce cargo damage, enabling modern logistics and warehousing to be mechanized and automated, thereby greatly reducing labor costs and lowering corporate logistics and warehousing expenses.

[0003] Based on the above advantages, pallets are widely used in logistics warehousing and supply chain industries, but there are still shortcomings in three aspects:

[0004] (1) Traditional manpower or forklifts are used to load and unload goods. After the goods are unloaded from the pallet, they are collected from the site by manpower or forklifts for stacking, which causes the loss of manpower and material resources of the enterprise, increases costs, and the stacking is not neat and the pallet stacking is skewed;

[0005] (2) Modern fully automated unmanned warehouses mainly realize the loading and unloading of goods on shelves and in and out of the warehouse. The pallets of fully automated unmanned warehouses are generally equipped with RFID tags and have specific specifications that match the shelves. They are generally not involved in market circulation. Automatic storage equipment such as AGVs and chain conveyors only deliver pallets with goods to the outbound point. At the back end, manpower or personnel driving forklifts will load the goods into trucks and containers. After loading, the collection and stacking of empty pallets left at the work site still rely on manpower or personnel driving forklifts, which also have the same problems as the first point.

[0006] (3) When loading goods onto trucks or containers, personnel must not only carry or operate equipment such as forklifts, but also pay attention to on-site safety. They must also count the goods loaded onto the trucks to check whether there are too many or too few goods loaded. Goods that have already been loaded onto trucks or containers are often unloaded and recounted, resulting in a waste of human and material resources and increased costs for the company.

[0007] by Figure 6 Taking the actual situation of the logistics and warehousing industry as an example, the above problems are demonstrated.

[0008] Factory Introduction: A chemical factory produces ABS granules. The finished products are delivered to the warehouse in 25 kg / bag, 40 bags / pallet, or 1 ton / pallet. When shipped out of the warehouse, they are packed and loaded with a rotary forklift. The pallets are circulated within the factory warehouse and are not shipped with the goods. The basic delivery process is as follows: Figure 4 Introduction.

[0009] Among them, the fourth problem is: because the empty pallet is light, when the forklift places the empty pallet and moves back, the pallet will move with the fork teeth, making it difficult to stack the pallets neatly; at the same time, the rotary pusher affects the forklift driver's vision, which is not conducive to the forklift driver stacking the pallets neatly. It takes about 15 seconds to place each pallet, and about 27 empty pallets are needed to load one container, which takes about 7 minutes in total.

[0010] ⑤ The fifth problem: The flat-fork forklift restacks and reorganizes the crooked empty pallets. Each truck does this three times, each time taking about three minutes, for a total of nine minutes.

[0011] The sixth problem: the forklift stacks the pallets in an untidy and skewed manner, and the deviation of the pallet stacking exceeds 40CM, which poses a safety hazard. The stacking heights are inconsistent, with a difference of more than 5 pieces. This is not conducive to the inventory management of empty pallets, and is even less conducive to the subsequent retrieval of empty pallets, especially not suitable for automatic retrieval in automatic warehouses, and needs to be sorted again.

[0012] Let's take a look at the container loading situation. Brief introduction: Containers have load weight limits and load balance requirements, so a reasonable layout of loading is required. Take the 45G1 container as an example. The cargo weight limit is 27,000KG, the cargo weight is 1,000KG / piece, and 27 pieces can be loaded.

[0013] The forklift truck uses a pallet to transport goods into the container, unloads the goods by rotating and pushing, and then takes the empty pallet out of the container. It takes 54 minutes for a forklift driver to load one container, and 27 minutes to go in and out of the container. The on-site operation environment is complex, and the driver's attention is mainly on the safety of the operation. He often forgets how much goods are loaded in the container, with an error rate of 2%. If an abnormality is found during the factory weighing, it will be returned to the warehouse for processing, and the forklift driver needs to crawl into the narrow space to count again, which wastes time. Summary of the invention

[0014] In order to solve the above technical problems, the present invention proposes a pallet automatic detection, counting and stacking device and a warehouse supporting automated process, which can ensure that the pallet is accurately aligned with the stacking position during the sliding process, greatly improves the accuracy of stacking, reduces the problem of unstable pallet placement caused by improper stacking, and monitors the pallet status, quantity and other information in real time, which is convenient for enterprises to make scientific decisions and optimize resource allocation.

[0015] To achieve the above object, the technical solution adopted by the present invention is:

[0016] A pallet automatic detection, counting and stacking device, characterized in that: it includes a pallet placement box for placing pallets, the upper end surface of the pallet placement box is provided with a delivery port for placing pallets, and the two opposite side surfaces of the pallet placement box are respectively provided with a visual statistics component and a pallet deceleration button, the visual statistics component includes a visual camera and a controller, the visual camera is connected to the controller, the controller is provided with a transmission interface for connecting to a central control system and a power supply interface for power supply, the pallet deceleration button includes a deceleration ball, a spring and an adjustment seat, the side of the pallet placement box is provided with a groove, and the adjustment seat is movable Installed in the groove, the spring is installed between the deceleration ball and the adjustment seat and its two ends are respectively fixed to the deceleration ball and the adjustment seat. An infrared sensor connected to the visual camera signal is arranged in the deceleration ball, and it also includes a reader and a counting display connected to the controller through data lines. An electronic identification code is installed on the pallet. When the pallet is put into or taken out of the pallet placement box, the pallet touches the deceleration ball, and the visual camera takes a picture of it to detect the intact state of the pallet and whether there is any remaining goods on the pallet. At this time, the reader reads the electronic identification code on the pallet, and the counting display automatically increases or decreases the count value.

[0017] In the above structure: the present invention proposes an automatic pallet detection, counting and stacking device, which includes a pallet placement box for placing pallets, and a visual counting component and a pallet deceleration button are respectively arranged on the two opposite sides of the pallet placement box. The pallet deceleration button can not only decelerate the falling pallet, but also trigger the visual detection component to take pictures to detect the intact state of the pallet and whether there is any remaining goods on the pallet.

[0018] Among them, the pallet deceleration button includes a deceleration ball, a spring and an adjustment seat. A groove is arranged on the side of the pallet placement box, and the adjustment seat is movably installed in the groove. The spring is installed between the deceleration ball and the adjustment seat and its two ends are respectively fixed to the deceleration ball and the adjustment seat. Therefore, when the adjustment seat moves, it will drive the deceleration ball to move together, thereby adjusting the extension length of the deceleration ball, so that the device can correspond to pallets of different specifications. An infrared sensor connected to the visual camera signal is arranged in the deceleration ball. The infrared sensor is used to detect whether the deceleration ball is touched. If the deceleration ball is touched, the infrared sensor inside it transmits a signal to the visual camera, and the visual camera starts shooting immediately. The deceleration ball will automatically pop out by the spring after the pallet passes.

[0019] The visual statistics component includes a visual camera and a controller. The visual camera is connected to the controller. The controller is provided with a transmission interface for connecting to the central control system and a power supply interface for power supply. The reader can read the electronic identification code information on the pallet in real time, accurately record and count the number of pallets, and realize the system to review the delivery quantity. When the pallet falls down, the deceleration ball is triggered and the visual camera starts shooting. The visual camera uses the image processing algorithm to perform real-time analysis on the captured image. The central control system can quickly identify the outline, shape and feature points of the empty pallet, and compare it with the preset pallet model to detect the intact state of the pallet and whether there is any remaining goods on the pallet, and determine the position and posture of the pallet. The transmission interface on the controller is connected to the central control system for data transmission, and the power supply interface is used to provide power for it. The controller can integrate and analyze the data, integrate the AGV transportation data, the visual camera recognition data and the reader reading data into the central control system for analysis and processing, and provide data support for production management.

[0020] As a preferred technical solution of the present invention: an upper baffle is installed on the delivery port of the pallet placement box, and the upper baffle is fixed and tilted outward.

[0021] In the above structure: the side baffles on the delivery port are fixed and tilted outward. After the AGV transports the empty pallet to the delivery port, the upper baffle, the pallet deceleration button and the side baffles adjust the falling route of the pallet, guiding the pallet to slide smoothly down to the stacking position to achieve automatic and neat stacking.

[0022] As a preferred technical solution of the present invention: the front end face of the pallet placement box is provided with a side opening, and side baffles are installed at the side opening, and the side baffles are respectively fixed on the two side faces of the pallet placement box. When the pallet is placed in the pallet placement box, the electronic identification code on the pallet is located between the side baffles.

[0023] In the above structure: the side opening is not only convenient for the staff to take out the tray, but also convenient for recording the electronic identification code on the tray. The side baffle installed on the side opening can prevent the tray from slipping.

[0024] As a preferred technical solution of the present invention: a universal wheel assembly is installed on the bottom surface of the pallet placement box, and the universal wheel assembly includes a universal wheel and a brake for braking and fixing the universal wheel.

[0025] In the above structure: the universal wheels installed at the bottom are convenient for the staff to push the pallet to place the box and move it, and the brake can brake and fix the universal wheels to prevent the car from slipping.

[0026] As a preferred technical solution of the present invention: a through hole is provided at the bottom of the groove, the adjustment seat is a T-shaped structure, the spring is fixedly connected to the upper end surface of the T-shaped adjustment seat, and the other end is arranged through the through hole and a nut is installed thereon.

[0027] In the above structure: the movement of the adjustment seat in the groove is manually adjusted by the staff. When the position of the deceleration ball needs to be adjusted, the nut at the other end of the adjustment seat is pulled to drive the deceleration ball to move with it, and then the nut is tightened forward or loosened backward to achieve the position adjustment of the deceleration ball.

[0028] As a preferred technical solution of the present invention: it also includes a camera protection cover, which is fixed on the side of the tray placement box and is located above the visual camera.

[0029] In the above structure: the camera protective cover is used to protect the visual camera and prevent damage.

[0030] An automated warehouse matching process for a pallet automatic detection, counting and stacking device, the specific steps are as follows, and the characteristics are as follows: Step 1: deploying AGV and manufacturing stacking device, introducing AGV into the production line, responsible for automatically transporting empty pallets to the stacking device delivery port, and at the same time modifying the pallet, adding RFID function, to ensure that each pallet has a unique electronic identification code;

[0031] Step 2: Install a machine vision system. Install a visual camera and a machine vision recognition system at the delivery port of the stacking device to identify and locate the delivered empty pallets, detect the intact state of the pallets and whether there is any remaining goods on the pallets;

[0032] Step 3: Automatic pallet stacking. After the AGV transports the empty pallet to the delivery port, the upper baffle, pallet deceleration button and side baffle adjust the pallet's falling route, guiding the pallet to slide smoothly down to the stacking position, achieving automatic and neat stacking. This function is also applicable to non-automatic warehouse enterprises without AGV and RFID.

[0033] Step 4: Optimize the counting and review process. Use the counting display to count the pallets in real time and optimize the steps of the delivery personnel to count the pallets. This function is also applicable to non-automatic warehouse enterprises without AGV.

[0034] Step 5: Read and record RFID information. Add a reader / writer to the pallet collector to read the RFID information on the pallet in real time, accurately record and count the number of pallets and the name, material number, quantity, batch and other information of the original goods on the pallet, realize automatic warehouse picking, destocking, ERP system IM posting and transportation plan delivery quantity, and conduct whole chain system review;

[0035] Step 6: Real-time monitoring and adjustment: The quality of the recovered empty pallets is monitored in real time through the machine vision system. The operator can remotely control the AGV to adjust the strategy of placing and removing pallets according to the actual situation to ensure the stable operation of the production line.

[0036] Step 7: Data integration and analysis: Integrate AGV transportation data, machine vision recognition data, and RFID reading data into the central control system for analysis and processing to provide data support for production management.

[0037] As a preferred technical solution of the present invention: the installation of the machine vision system in step 2 is represented as follows:

[0038] The installation of the machine vision system in step 2 is as follows:

[0039] Step 2.1, install a visual camera at the delivery port of the stacking device. The visual camera has high resolution and can clearly capture the image details of the empty pallet. At the same time, it can ensure that high-quality images can be stably obtained under different lighting conditions. Even in a dark environment, the features of the pallet can be accurately captured.

[0040] Step 2.2, establish a machine vision recognition system, and use image processing algorithms to perform real-time analysis on the images captured by the visual camera. The system can quickly identify the outline, shape, and feature points of the empty pallet, and compare it with the preset pallet model to detect the intact state of the pallet and whether there is any remaining goods on the pallet, and determine the position and posture of the pallet;

[0041] Step 2.3, after identifying the pallet, the system calculates the pallet's center position, angle and other parameters to provide accurate data for subsequent guidance operations; during the pallet's descent, after the AGV transports the empty pallet to the delivery port, the upper baffle, pallet deceleration button and side baffle adjust the pallet's falling route, guiding the pallet to slide smoothly to the stacking position for automatic and neat stacking.

[0042] As a preferred technical solution of the present invention: Step 2.2, establishing a machine vision recognition system is expressed as follows:

[0043] Step 2.2.1, image acquisition and preprocessing, use a visual camera to acquire images at the stacking device delivery port at a fixed frame rate, and perform preprocessing operations on the images:

[0044] I new (x,y)=Median{I(x+i,y+j)}

[0045] Among them, I new (x, y) is the pixel value of the denoised image, I(x, y) is the pixel value of the original image, (x, y) is the image coordinate point, Median is the median function, i, j∈[-n, n], n is the filter window size;

[0046] Step 2.2.2, pallet contour extraction, the pallet contour is extracted by edge detection algorithm, and the extracted gradient amplitude and gradient direction are expressed as:

[0047]

[0048] Where G is the gradient amplitude of the pallet profile, θ is the gradient direction of the pallet profile, and G x and G y are the gradients in the horizontal and vertical directions respectively;

[0049] Step 2.2.3, shape and feature point recognition, for the shape recognition of the tray, the shape descriptor Hu moment is used:

[0050] φ1=η 20 +η 02

[0051]

[0052] φ3=(η 30 -3η 12 ) 2 +(3η 21 -η 03 ) 2

[0053] Among them, φ1, φ2, φ3 are the invariant moments of Hu moments, η 20 , η 02 , η 11 , η 30 , η 12 , η 12 , η 21 , η 03 It is an intermediate variable when calculating the Hu moment. The subscripts represent different powers, which are obtained by weighted summing the image function at different coordinate points with different powers.

[0054] Step 2.2.4, compare with the preset model, compare the extracted pallet contour, shape features and feature points with the preset pallet model, the comparison formula is as follows:

[0055]

[0056] Among them, T(x,y) is the template image, I(x,y) is the pixel value of the input image at the coordinate point (x,y), and are the means of the template image and the input image, respectively, and NCC is the normalized cross-correlation value, which is used to measure the similarity between the template image and the input image, and quickly and accurately determine whether the object in the current image is the target pallet by comparing with the preset model;

[0057] Step 2.2.5, calculate the center position and angle. The center position of the pallet is determined by calculating the centroid of the pallet contour. The formula is as follows:

[0058]

[0059] Among them, x c and c Respectively represent the coordinates of the center of mass of the pallet contour in the horizontal and vertical directions, (x i ,y i ) is the coordinate of the pixel point on the contour, and n is the total number of pixel points on the pallet contour;

[0060] The covariance matrix of the pallet contour points is calculated, and then the eigenvectors and eigenvalues ​​of the covariance matrix are obtained. The direction of the eigenvector corresponding to the larger eigenvalue is the main direction of the pallet, thereby determining the angle of the pallet.

[0061] As a preferred technical solution of the present invention: Step 3 guides the tray to slide smoothly down to the stacking position as follows:

[0062] Step 3.1, during the sliding process of the pallet, the motion state of the pallet is monitored by continuously acquiring images and analyzing the position change of the pallet between different frames;

[0063] Step 3.2, according to the movement state of the pallet, adjust the guidance strategy in real time. If the pallet deviates from the predetermined stacking position, the angle, speed and other parameters of the stacking device can be adjusted to guide the pallet back to the correct position. The adjusted guiding force can be expressed as:

[0064] F adjusted =F original +k×Δd

[0065] Among them, F original is the original guiding force, k is the adjustment coefficient, Δd is the offset between the pallet and the target position, and F adjusted It is the adjusted guiding force.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] 1. The visual camera in the present invention uses an image processing algorithm to perform real-time analysis on the captured images. The central control system can quickly identify the outline, shape and feature points of the empty pallet, compare it with the preset pallet model, perform quality inspection on the pallet and whether there is any remaining goods on the pallet, determine the position and posture of the pallet, adjust the falling route of the pallet through the upper baffle, pallet deceleration button and side baffle, and guide the pallet to slide down smoothly to the stacking position, which greatly improves the accuracy of stacking and reduces the problems of unstable pallet placement caused by improper stacking.

[0068] 2. The present invention sets a pallet deceleration button. The deceleration ball slows down the falling speed of the pallet and sends a signal through the infrared sensor therein to trigger the visual camera to take pictures and detect. The deceleration ball automatically pops out through a spring after passing through. The depth of the deceleration ball can also be adjusted through an adjustment seat, so that the device can correspond to pallets of different specifications, improves the neatness of pallet stacking and the uniformity of the number of empty pallets in the AGV fork collection device, and reduces problems such as inconsistent stacking quantity.

[0069] 3. The present invention integrates and analyzes data through the controller, integrates AGV transportation data, visual camera recognition data, and reader-writer reading data into the central control system for analysis and processing, provides data support for production management, and realizes intelligent management of pallets. It can monitor the status and quantity of pallets in real time, which is convenient for enterprises to make scientific decisions and optimize resource allocation.

[0070] 4. The present invention uses AGV to automatically transport empty pallets, and the machine vision system quickly identifies and locates pallets, reducing manual lifting and stacking operations, saving time and labor costs, and significantly improving the efficiency of the entire production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a schematic diagram of the structure of the automatic detection, counting and stacking device for pallets in the present invention;

[0072] Figure 2 It is a schematic diagram of the installation of the visual statistics component in the present invention;

[0073] Figure 3 This is a schematic diagram of the installation of the tray slow-drop button in the present invention;

[0074] Figure 4 It is a flow chart of the warehouse supporting automation process in the present invention;

[0075] Figure 5 It is a practical business scenario diagram of the present invention;

[0076] Figure 6 It is the basic delivery process in the logistics and warehousing industry.

[0077] List of reference numerals:

[0078] 1. Pallet; 2. Pallet placement box; 3. Electronic identification code; 4. Reader / writer; 5. Counting display; 6. Pallet slow-down button; 61. Slow-down ball; 62. Spring; 63. Adjustment seat; 7. Camera protective cover; 8. Visual camera; 9. Controller; 10. Transmission interface; 11. Power interface; 12. Universal wheel assembly; 13. Upper baffle plate; 14. Side baffle plate; 15. Groove. DETAILED DESCRIPTION

[0079] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0080] like Figure 1-3 As shown, the present invention proposes an automatic detection, counting and stacking device for pallets, comprising a pallet placement box 2 for placing pallets 1, the upper end surface of the pallet placement box 2 is provided with a delivery port for placing the pallet 1, and the two opposite sides of the pallet placement box 2 are respectively provided with a visual statistics component and a pallet deceleration button 6, the visual statistics component comprises a visual camera 8 and a controller 9, the visual camera 8 is connected to the controller 9, and the controller 9 is provided with a transmission interface 10 for connecting to a central control system and a power supply interface 11 for power supply, the pallet deceleration button 6 comprises a deceleration ball 61, a spring 62 and an adjustment seat 63, and a groove 15 is provided on the side of the pallet placement box 2. The adjustment seat 63 is movably installed in the groove 15, the spring 62 is installed between the deceleration ball 61 and the adjustment seat 63 and its two ends are respectively fixed to the deceleration ball 61 and the adjustment seat 63, an infrared sensor connected to the signal of the visual camera 8 is arranged in the deceleration ball 61, and also includes a reader 4 and a counting display 5 connected to the controller 9 through data lines respectively, an electronic identification code 3 is installed on the pallet 1, when the pallet 1 is placed in or taken out of the pallet placement box 2, the pallet 1 touches the deceleration ball 61, and the visual camera 8 takes a picture of it. At this time, the reader 4 reads the electronic identification code 3 on the pallet 1, and the counting display 5 automatically increases or decreases the count value.

[0081] An upper baffle 13 is installed on the loading port of the pallet placement box 2, and the upper baffle 13 is fixed and tilted outward.

[0082] The front end face of the pallet placement box 2 is provided with a side opening, and a side baffle 14 is installed at the side opening. The side baffles 14 are respectively fixed on the two side faces of the pallet placement box 2. When the pallet 1 is placed in the pallet placement box 2, the electronic identification code 3 on the pallet 1 is located between the side baffles 14.

[0083] A universal wheel assembly 12 is installed on the bottom surface of the pallet placement box 2. The universal wheel assembly 12 includes a universal wheel and a brake for braking and fixing the universal wheel.

[0084] A through hole is provided at the bottom of the groove 15, the adjustment seat 63 is a T-shaped structure, the spring 62 is fixedly connected to the upper end surface of the T-shaped adjustment seat 63, and the other end is arranged through the through hole and a nut is installed thereon.

[0085] A camera protection cover 7 is also included, which is fixed on the side of the tray placement box 2 and is located above the visual camera 8.

[0086] The present invention proposes an automatic pallet detection, counting and stacking device, which includes a pallet placement box 2 for placing pallets 1. Visual statistical components and pallet deceleration buttons 6 are respectively arranged on two opposite sides of the pallet placement box 2. The visual statistical components can count the data of the pallets 1 stored in the pallet placement box 2. The pallet 1 deceleration button can not only decelerate the falling pallet 1, but also trigger the visual statistical component to take pictures.

[0087] Among them, the tray deceleration button 6 includes a deceleration ball 61, a spring 62 and an adjustment seat 63. A groove 15 is arranged on the side of the tray placement box 2. The adjustment seat 63 is movably installed in the groove 15. The spring 62 is installed between the deceleration ball 61 and the adjustment seat 63 and its two ends are respectively fixed to the deceleration ball 61 and the adjustment seat 63. Therefore, when the adjustment seat 63 moves, it will drive the deceleration ball 61 to move together, thereby adjusting the extension length of the deceleration ball 61, so that the device can correspond to pallets 1 of different specifications. An infrared sensor connected to the signal of the visual camera 8 is arranged in the deceleration ball 61. The infrared sensor is used to detect whether the deceleration ball 61 is touched. If the deceleration ball 61 is touched, the infrared sensor therein transmits a signal to the visual camera 8, and the visual camera 8 starts shooting immediately. The deceleration ball 61 will automatically pop out through the spring 62 after the tray 1 passes.

[0088] The visual statistics component includes a visual camera 8 and a controller 9. The visual camera 8 is connected to the controller 9. The controller 9 is provided with a transmission interface 10 for connecting to the central control system and a power supply interface 11 for power supply. The reader 4 can read the electronic identification code information on the pallet 1 in real time, accurately record and count the number of pallets 1, and realize the system to review the delivery quantity. When the pallet 1 falls down, the deceleration ball 61 is triggered, and the visual camera 8 starts shooting. The visual camera 8 uses an image processing algorithm to perform real-time analysis on the captured image. The central control system can quickly identify the outline, shape and feature points of the empty pallet, and determine the position and posture of the pallet 1 by comparing it with the preset pallet model. The transmission interface 10 on the controller 9 is connected to the central control system for data transmission, and the power supply interface 11 is used to provide it with power. The controller 9 can integrate and analyze the data, integrate the AGV transportation data, the visual camera 8 recognition data and the reader 4 reading data into the central control system for analysis and processing, and provide data support for production management.

[0089] In this embodiment, the side baffles 14 on the delivery port are fixed and tilted outward. After the AGV transports the empty pallet to the delivery port, the upper baffle 13, the pallet deceleration button 6 and the side baffles 14 adjust the falling route of the pallet 1, guiding the pallet 1 to slide smoothly down to the stacking position to achieve automatic and neat stacking.

[0090] In this embodiment, the side opening is not only convenient for taking out the tray 1, but also convenient for recording the electronic identification code 3 on the tray 1. The side baffle 14 installed on the side opening can prevent the placed tray 1 from sliding off.

[0091] In this embodiment, the universal wheels installed at the bottom are convenient for the staff to push the tray placement box 2 to move, and the brake can brake and fix the universal wheels to prevent the vehicle from slipping.

[0092] In this embodiment, the movement of the adjustment seat 63 in the groove 15 is manually adjusted by the staff. When the position of the deceleration ball 61 needs to be adjusted, the nut at the other end of the adjustment seat 63 is pulled to drive the deceleration ball 61 to move therewith, and then the nut is tightened forward or loosened backward to achieve the position adjustment of the deceleration ball 61.

[0093] In this embodiment, the camera protection cover 7 is used to protect the visual camera 8 to prevent damage.

[0094] like Figure 4-5 As shown, the present invention proposes a warehouse matching automation process for a pallet automatic detection, counting and stacking device, and the specific steps are as follows:

[0095] Step 1: Deploy AGV and manufacture stacking devices. Introduce AGV into the production line to automatically transport empty pallets to the stacking device delivery port. At the same time, modify pallet 1 and add RFID function to ensure that each pallet 1 has a unique electronic identification code 3;

[0096] Step 2: Install a machine vision system. Install a visual camera 8 and a machine vision recognition system at the delivery port of the stacking device to identify and locate the delivered empty pallets, and detect the intact state of the pallets and whether there are any remaining goods on the pallets;

[0097] Step 3: Automatically stack the pallet 1. After the AGV transports the empty pallet to the delivery port, the upper baffle 13, the pallet deceleration button 6 and the side baffle 14 adjust the falling route of the pallet 1, guide the pallet 1 to slide smoothly down to the stacking position, and realize automatic and neat stacking. This function is also applicable to non-automatic warehouse enterprises without AGV and RFID.

[0098] Step 4: Optimize the counting and review process, use the counting display 5 to count the pallet 1 in real time, and optimize the step of the delivery personnel counting the pallets. This function is also applicable to non-automatic warehouse enterprises without AGV;

[0099] Step 5: Read and record RFID information. Add a reader / writer 4 to the collector of pallet 1 to read the RFID information on pallet 1 in real time, accurately record and count the number of pallets 1 and the name, material number, quantity, batch and other information of the original goods on the pallet, realize automatic warehouse picking, delisting, ERP system IM posting and the delivery quantity of the transportation plan, and conduct a review of the entire chain system;

[0100] Step 6: Real-time monitoring and adjustment: The quality of the recovered empty pallets is monitored in real time through the machine vision system. The operator can remotely control the AGV to adjust the strategy of placing and removing pallets according to the actual situation to ensure the stable operation of the production line.

[0101] Step 7: Data integration and analysis: Integrate AGV transportation data, machine vision recognition data, and RFID reading data into the central control system for analysis and processing to provide data support for production management.

[0102] The installation of the machine vision system in step 2 is as follows:

[0103] Step 2.1, a visual camera 8 is installed at the delivery port of the stacking device. The visual camera 8 has high resolution and can clearly capture the image details of the delivered empty pallet. At the same time, it can ensure that high-quality images can be stably obtained under different lighting conditions. Even in a dark environment, the features of the pallet 1 can be accurately captured;

[0104] Step 2.2, establish a machine vision recognition system, and use image processing algorithms to perform real-time analysis on the images captured by the visual camera 8. The system can quickly identify the outline, shape and feature points of the empty pallet, and compare it with the preset pallet model to detect the intact state of the pallet and whether there is any remaining goods on the pallet, and determine the position and posture of the pallet 1;

[0105] Step 2.3, after identifying pallet 1, the system calculates the center position, angle and other parameters of pallet 1 to provide accurate data for subsequent guidance operations; during the sliding process of pallet 1, after the AGV transports the empty pallet to the delivery port, the upper baffle 13, pallet deceleration button 6 and side baffle 14 adjust the falling route of pallet 1, and guide pallet 1 to slide smoothly to the stacking position to achieve automatic and neat stacking.

[0106] Step 2.2, establishing a machine vision recognition system is as follows:

[0107] Step 2.2.1, image acquisition and preprocessing, using the visual camera 8 to acquire the image at the stacking device delivery port at a fixed frame rate, and performing preprocessing operations on the image:

[0108] I new (x,y)=Median{I(x+i,y+j)}

[0109] Among them, I new (x, y) is the pixel value of the denoised image, I(x, y) is the pixel value of the original image, (x, y) is the image coordinate point, Median is the median function, i, j∈[-n, n], n is the filter window size;

[0110] Step 2.2.2, Pallet 1 contour extraction, the contour of pallet 1 is extracted by edge detection algorithm, and the extracted gradient amplitude and gradient direction are expressed as:

[0111]

[0112] Where G is the gradient amplitude of the contour of tray 1, θ is the gradient direction of the contour of tray 1, and G x and G y are the gradients in the horizontal and vertical directions respectively;

[0113] Step 2.2.3, shape and feature point recognition, for the shape recognition of tray 1, the shape descriptor Hu moment is used:

[0114] φ1=η 20 +η 02

[0115]

[0116] φ3=(η 30 -3η 12 ) 2 +(3η 21 -η 03 ) 2

[0117] Among them, φ1, φ2, φ3 are the invariant moments of Hu moments, η 20 , η 02 , η 11 , η 30 , η 12 , η 12 , η 21 , η 03 It is an intermediate variable when calculating the Hu moment. The subscripts represent different powers, which are obtained by weighted summing the image function at different coordinate points with different powers.

[0118] Step 2.2.4, compare with the preset model, compare the extracted pallet 1 contour, shape features and feature points with the preset pallet model, the comparison formula is as follows:

[0119]

[0120] Among them, T(x,y) is the template image, I(x,y) is the pixel value of the input image at the coordinate point (x,y), and are the means of the template image and the input image, respectively. NCC is the normalized cross-correlation value, which is used to measure the similarity between the template image and the input image. By comparing with the preset model, it is quickly and accurately determined whether the object in the current image is the target tray 1;

[0121] Step 2.2.5, calculate the center position and angle. The center position of tray 1 is determined by calculating the centroid of the contour of tray 1. The formula is as follows:

[0122]

[0123] Among them, x c and c Respectively represent the coordinates of the center of mass of the pallet contour in the horizontal and vertical directions, (x i ,y i ) is the coordinate of the pixel point on the contour, and n is the total number of pixel points on the contour of tray 1;

[0124] The covariance matrix of the contour points of tray 1 is calculated, and then the eigenvector and eigenvalue of the covariance matrix are obtained. The direction of the eigenvector corresponding to the larger eigenvalue is the main direction of tray 1, thereby determining the angle of tray 1.

[0125] Step 3 guides tray 1 to slide down smoothly to the stacking position as shown below:

[0126] Step 3.1, during the sliding process of the tray 1, the movement state of the tray 1 is monitored by continuously acquiring images and analyzing the position change of the tray 1 between different frames;

[0127] Step 3.2, according to the motion state of tray 1, adjust the guidance strategy in real time. If tray 1 deviates from the predetermined stacking position, the angle, speed and other parameters of the stacking device can be adjusted to guide tray 1 back to the correct position. The adjusted guidance force can be expressed as:

[0128] F adjusted =F original +k×Δd

[0129] Among them, F original is the original guiding force, k is the adjustment coefficient, Δd is the offset between tray 1 and the target position,

[0130] F adjusted It is the adjusted guiding force.

[0131] 1. The visual camera 8 of the present invention uses an image processing algorithm to perform real-time analysis on the captured images. The central control system can quickly identify the outline, shape and feature points of the empty pallet, and determine the position and posture of the pallet 1 by comparing it with the preset pallet model. The upper baffle 13, the pallet deceleration button 6 and the side baffle 14 are used to adjust the falling route of the pallet 1, and guide the pallet 1 to slide smoothly down to the stacking position, which greatly improves the accuracy of stacking and reduces the problem of unstable placement of the pallet 1 caused by improper stacking.

[0132] 2. The present invention sets a tray deceleration button 6. The deceleration ball 61 decelerates the falling speed of the tray 1 and sends a signal through the infrared sensor therein, triggering the visual camera 8 to take pictures and detect and the counting display 5 to count changes. After the deceleration ball 61 passes, it is automatically ejected by the spring 62. The depth of the deceleration ball 61 can also be adjusted by the adjustment seat 63, so that the device can correspond to trays 1 of different specifications.

[0133] 3. The present invention integrates and analyzes data through the central control system, integrates AGV transportation data, visual camera 8 recognition data, and reader 4 reading data into the central control system for analysis and processing, provides data support for production management, and realizes intelligent management of pallets 1. The status, quantity and other information of pallets 1 can be monitored in real time, which is convenient for enterprises to make scientific decisions and optimize resource allocation.

[0134] 4. The present invention uses AGV to automatically transport empty pallets, and the machine vision system quickly identifies and locates the pallet 1, reducing manual lifting and stacking operations, saving time and labor costs, and significantly improving the efficiency of the entire production process.

[0135] The above description is only a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent change made based on the technical essence of the present invention still falls within the scope of protection required by the present invention.

Claims

1. A device for automatically detecting, counting and stacking pallets, characterized in that: The invention comprises a pallet placement box (2) for placing a pallet (1), the upper end surface of the pallet placement box (2) being provided with a delivery port for placing the pallet (1), and two opposite side surfaces of the pallet placement box (2) being provided with a visual statistics component and a pallet deceleration button (6), respectively, the visual statistics component comprising a visual camera (8) and a controller (9), the visual camera (8) being connected to the controller (9), and the controller (9) being provided with a transmission interface (10) for connecting to a central control system and a power supply interface (11) for power supply. 11), the tray deceleration button (6) comprises a deceleration ball (61), a spring (62) and an adjustment seat (63), a groove (15) is arranged on the side of the tray placement box (2), the adjustment seat (63) is movably installed in the groove (15), the spring (62) is installed between the deceleration ball (61) and the adjustment seat (63) and its two ends are respectively fixed to the deceleration ball (61) and the adjustment seat (63), an infrared sensor connected to the signal of the visual camera (8) is arranged in the deceleration ball (61), and the infrared sensor The sensor is used to detect whether the deceleration ball (61) is touched. If the deceleration ball (61) is touched, the infrared sensor therein transmits a signal to the visual camera (8), and the visual camera (8) immediately starts shooting. The sensor also includes a reader (4) and a counting display (5) respectively connected to the controller (9) through data lines. An electronic identification code (3) is installed on the tray (1). When the tray (1) is placed in or taken out of the tray placement box (2), the tray (1) touches the deceleration ball (61), and the visual camera (8) captures the deceleration ball (61). Taking pictures, detecting the state of the pallet through a transmission interface (10) connected to the central control system by the controller (9), the visual camera (8) performing real-time analysis on the taken images through an image processing algorithm, the central control system can quickly identify the outline, shape and feature points of the empty pallet (1), and determine the position and posture of the pallet (1) by comparing with a preset pallet model, the reader (4) reading the electronic identification code (3) on the pallet (1), and the counting display (5) automatically increasing or decreasing the count value; An upper baffle (13) is installed on the loading port of the tray placement box (2), and the upper baffle (13) is tilted and fixed outwards; The front end surface of the pallet placement box (2) is provided with a side opening, and a side baffle (14) is installed at the side opening. The side baffles (14) are respectively fixed on the two side surfaces of the pallet placement box (2). When the pallet (1) is placed in the pallet placement box (2), the electronic identification code (3) on the pallet (1) is located between the side baffles (14). The upper baffle (13), the pallet deceleration button (6) and the side baffles (14) adjust the falling route of the pallet (1) to guide the pallet (1) to slide smoothly to the stacking position. If the pallet (1) deviates from the predetermined stacking position, the angle and speed parameters of the stacking device are adjusted to guide the pallet (1) back to the correct position.

2. The automatic detection, counting and stacking device for pallets according to claim 1, characterized in that: A universal wheel assembly (12) is installed on the bottom surface of the pallet placement box (2), and the universal wheel assembly (12) comprises a universal wheel and a brake for braking and fixing the universal wheel.

3. The automatic detection, counting and stacking device for pallets according to claim 1, characterized in that: A through hole is provided at the bottom of the groove (15); the adjustment seat (63) is a T-shaped structure; the spring (62) is fixedly connected to the upper end surface of the T-shaped adjustment seat (63); the other end passes through the through hole and a nut is installed thereon.

4. The automatic detection, counting and stacking device for pallets according to claim 1, characterized in that: It also includes a camera protection cover (7), which is fixed on the side of the pallet placement box (2) and is located above the visual camera (8).

5. A method for using the automatic pallet detection, counting and stacking device in a warehouse according to any one of claims 1 to 4, wherein the specific steps are as follows, and the characteristics are as follows: Step 1: Deploy AGV and manufacture stacking devices. Introduce AGV into the production line to automatically transport empty pallets (1) to the stacking device delivery port. At the same time, modify the pallets (1) and add RFID functions to ensure that each pallet (1) has a unique electronic identification code (3); Step 2: Install a machine vision system, install a visual camera (8) and a machine vision recognition system at the delivery port of the stacking device to identify and locate the delivered empty pallet (1), and detect the intact state of the pallet and whether there is any remaining goods on the pallet; Step 3: Automatically stack the pallet (1). After the AGV transports the empty pallet (1) to the delivery port, the upper baffle (13), the pallet deceleration button (6) and the side baffle (14) adjust the falling route of the pallet (1), guide the pallet (1) to slide smoothly down to the stacking position, and realize automatic and neat stacking. Step 4: Using the counting display (5), the tray (1) is dynamically counted in real time; Step 5: Reading and recording RFID information, adding a reader / writer (4) to the pallet (1) collector, reading the RFID information on the pallet (1) in real time, accurately recording and counting the number of pallets (1) and the name, material number, quantity, and batch information of the original cargo on the pallet; Step 6: Real-time monitoring and adjustment: The quality of the recovered empty pallets (1) is monitored in real time by the machine vision system. The operator can remotely control the AGV to adjust the strategy of placing and removing pallets according to the actual situation. Step 7: Data integration and analysis: Integrate AGV transportation data, machine vision recognition data, and RFID reading data into the central control system for analysis and processing.

6. A method for using the automatic pallet detection, counting and stacking device in a warehouse according to claim 5, characterized in that: The installation of the machine vision system in step 2 is as follows: Step 2.1, installing a visual camera (8) at the delivery port of the stacking device, wherein the visual camera (8) has a high resolution and can clearly capture image details of the delivered empty tray (1), while ensuring that high-quality images can be stably obtained under different lighting conditions; Step 2.2, establish a machine vision recognition system, and use an image processing algorithm to perform real-time analysis on the image captured by the visual camera (8). The system can quickly identify the outline, shape and feature points of the empty pallet (1), and by comparing it with a preset pallet model, detect the intact state of the pallet and whether there is any remaining goods on the pallet, and determine the position and posture of the pallet (1); Step 2.3, after identifying the tray (1), the system calculates the center position and angle parameters of the tray (1) to provide accurate data for subsequent guidance operations.

7. The warehouse supporting automation process of the pallet automatic detection, counting and stacking device according to claim 6 is characterized by: Step 2.2, establishing a machine vision recognition system is as follows: Step 2.2.1, image acquisition and preprocessing, using a visual camera (8) to acquire images at the stacking device delivery port at a fixed frame rate, and performing preprocessing operations on the images: ； in, is the pixel value of the denoised image, is the original image pixel value, is the image coordinate point, is the median function, , is the filter window size; Step 2.2.2, contour extraction of the tray (1), the contour of the tray (1) is extracted by edge detection algorithm, and the extracted gradient amplitude and gradient direction are expressed as: ； where G is the gradient amplitude of the contour of the tray (1), is the gradient direction of the tray (1) contour, and are the gradients in the horizontal and vertical directions respectively; Step 2.2.3, shape and feature point recognition, for the shape recognition of the tray (1), the shape descriptor Hu moment is used: ; in, is the invariant moment of the Hu moment, It is an intermediate variable when calculating Hu moment. The subscripts represent different powers. It is obtained by weighted summing the image function at different coordinate points with different powers. Step 2.2.4, compare with the preset model, compare the extracted pallet (1) contour, shape features and feature points with the preset pallet model, and the comparison formula is as follows: ; in, is the template image, is the input image at coordinate point The pixel value at and are the means of the template image and the input image, respectively. is the normalized cross-correlation value, which is used to measure the similarity between the template image and the input image, and to quickly and accurately determine whether the object in the current image is the target pallet by comparing it with the preset model (1); Step 2.2.5, calculate the center position and angle, and determine the center position of the tray (1) by calculating the centroid of the tray (1) contour. The formula is as follows: ; in, and Respectively represent the coordinates of the center of mass of the pallet contour in the horizontal and vertical directions, are the pixel coordinates on the contour, is the total number of pixels on the outline of the tray (1); The covariance matrix of the contour points of the pallet (1) is calculated, and then the eigenvector and eigenvalue of the covariance matrix are obtained. The direction of the eigenvector corresponding to the larger eigenvalue is the main direction of the pallet (1), thereby determining the angle of the pallet (1).

8. The warehouse supporting automation process of the pallet automatic detection, counting and stacking device according to claim 5 is characterized by: Step 3: Guide the tray (1) to slide down smoothly to the stacking position as shown below: Step 3.1, during the sliding process of the tray (1), the movement state of the tray (1) is monitored by continuously acquiring images and analyzing the position change of the tray (1) between different frames; Step 3.2, according to the motion state of the tray (1), the guiding strategy is adjusted in real time. If the tray (1) deviates from the predetermined stacking position, the angle and speed parameters of the stacking device are adjusted to guide the tray (1) back to the correct position. The adjusted guiding force can be expressed as: ; in, It is the original guiding force. is the adjustment factor, is the offset of the pallet (1) from the target position, It is the adjusted guiding force.

Images