A method, apparatus, and system for identifying the density of stacked plates

Abstract

This invention discloses a method, apparatus, and system for identifying the density of stacked boards, including a stacked board density identification system. The system comprises a material preparation platform for placing the stacked boards, as well as a thickness detection component, a weight detection component, and a length and width detection component. This invention incorporates the thickness detection component, weight detection component, and length and width detection component. Multiple distance sensors in the length and width detection component reduce errors in calculating length and width margins, while perspective transformation is used in the thickness detection component to reduce errors in calculating the thickness of individual boards. This addresses the problem of large density parameter calculation errors in existing fully automated board purification processes.

Description

Technical Field

[0001] This invention relates to the field of sheet metal processing technology, and specifically to a method, apparatus, and system for identifying the density of stacked sheet metal. Background Technology

[0002] In the automated purification process of wood panels, the panels need to be transported sequentially to a high-temperature formaldehyde removal chamber, a humidification chamber, and a spraying agent chamber for treatment. The high-temperature formaldehyde removal chamber heats the panels to accelerate formaldehyde release, the humidification chamber controls the humidity after the panels reach high temperatures, and the spraying agent chamber sprays agents with formaldehyde removal or odor neutralization functions onto the panel surface. For the automated purification process, initial state data of the panels to be purified needs to be obtained, including: material, density, thickness, length and width, moisture content, and initial environmental protection level (initial formaldehyde release). Based on the obtained initial state data, the initial purification mode related to parameters such as temperature and humidification is determined to achieve fully automated purification of the panels. However, during placement, the stacked panels often do not perfectly align with the central axis, causing the centerline of the stacked panels to be misaligned with the centerline of the preparation platform. This results in errors in length and width calculations. Furthermore, the tilted or distorted images captured during photography due to the shooting angle significantly reduce the accuracy of single-panel thickness identification, thus causing errors in single-panel density calculations. Summary of the Invention

[0003] The purpose of this invention is to provide a method, apparatus, and system for identifying the density of stacked boards. The method includes a thickness detection component, a weight detection component, and a length and width detection component. The length and width detection component uses multiple sets of distance sensors to reduce the error in calculating the length and width distances, and the thickness detection component uses a perspective transformation method to reduce the error in calculating the thickness of a single board. This invention aims to solve the problem of large errors in the calculation of density parameters in the existing fully automated purification process of boards.

[0004] A stacked board density identification system includes a preparation platform for placing stacked boards and a thickness detection component, a weight detection component, and a length and width detection component, wherein:

[0005] The thickness detection component includes a distance sensor group C and a camera component disposed above the material preparation platform, wherein the distance sensor group C is used to measure the height of the stacked plates, and the camera is used to acquire images of the side of the stacked plate blocks;

[0006] The length and width detection component includes distance sensor groups A and B. Distance sensor group A includes distance sensors respectively installed on the feeding side and the side opposite to the feeding side on the top surface of the material preparation platform. Distance sensor group B includes two distance sensors respectively installed on the top surface of the material preparation platform perpendicular to the feeding side. It also includes a height distance sensor installed above the top surface of the material preparation platform and a controller.

[0007] The weight detection component includes a weight sensor installed on the bottom surface of the material preparation platform;

[0008] It also includes a controller, which is connected to the thickness detection component, the weight detection component, and the length and width detection component respectively, to receive data from the thickness detection component, the weight detection component, and the length and width detection component and perform density calculation.

[0009] Furthermore, the distance sensor group C includes at least one distance sensor whose sensing direction is directed towards the top surface of the stacked plate block, for detecting the distance between the sensor and the stacked plate block in the height direction.

[0010] Furthermore, the camera assembly includes multiple cameras deployed along the height direction diagonally above the stack of boards, with each camera forming a different angle with the center of the top surface of the stacked boards, for capturing measurement images of the side of the stacked boards from different heights.

[0011] Furthermore, the distance sensor group A includes four distance sensors A1, A2, A3, and A4. Distance sensors A1 and A2 are located on the feeding side, and distance sensors A3 and A4 are located on the side opposite to the feeding side. The sensing directions of distance sensors A1, A2, A3, and A4 during measurement are respectively facing the two parallel sides of the stacked plate block.

[0012] Furthermore, the distance sensor group B includes four distance sensors B1, B2, B3, and B4. Each pair of distance sensors B1, B2, B3, and B4 is arranged in pairs. Each pair of distance sensors is respectively set on both sides of the top surface of the material preparation platform perpendicular to the feeding side, so that the sensing direction of the distance sensors A1 and A2 during measurement is directly facing the side of the stacked plate block.

[0013] Furthermore, each side of the top surface of the material preparation platform is provided with an embedded slide rail and two sliders set on the slide rail. Distance sensors A1, A2, A3, A4, B1, B2, B3, and B4 are respectively installed on the sliders.

[0014] A method for identifying the density of stacked boards specifically includes the following steps:

[0015] Receive data from the thickness detection component and perform thickness calculation to obtain the thickness h of the single board and the number N of single boards in the board stack block;

[0016] Receive data from the length and width detection components and perform length and width calculations to obtain the length L and width W of the single board;

[0017] Receive weight data M from the weight detection component;

[0018] The density of the veneer is calculated based on the thickness h of the veneer, the number N of veneers in the stacked veneer block, the length L of the veneer, the width W of the veneer, and the weight M.

[0019] Furthermore, the process of receiving data from the thickness detection component and performing thickness calculations to obtain the thickness h of the single board and the number N of single boards in the board stack block specifically includes the following steps:

[0020] S001. Obtain the first distance D0 between the distance sensor in the distance sensor group C and the top surface of the plate stack block;

[0021] S002. Based on the first preset distance DX0 and the first distance D0 between the distance sensor group and the material preparation platform, calculate the height H of the stacked plate block as H = DX1 - D0.

[0022] S003. Acquire the measurement image captured by the camera, and determine whether the measurement image contains the sides of all the single boards;

[0023] When it is determined that the measurement image contains the sides of all the single boards, the number of boundary lines K in the measurement image is obtained by using a pre-established recognition model;

[0024] The number of individual boards N in the board stack block is calculated based on the number of boundary lines K.

[0025] S004. Based on the height H of the stacked board block and the number of individual boards N, calculate the thickness h of the individual board corresponding to the stacked board block, where h = H / N.

[0026] Furthermore, the process of receiving data from the length and width detection component and performing length and width calculations to obtain the length L and width W of the board specifically includes the following steps:

[0027] S01. Acquire the data measured by distance haptic sensor group A and generate coordinate group AA; acquire the data measured by distance haptic sensor group B and generate coordinate group BB.

[0028] S02. Calculate the equations of the edge lines based on the coordinate set AA and coordinate set BB.

[0029] S03. Calculate the set of intersection points of the edge lines based on the set of edge line equations;

[0030] S04. Calculate the edge distance of the stacked board blocks based on the intersection point group, that is, the length L and width W of the single board.

[0031] A control device for identifying the density of stacked boards, characterized in that it comprises:

[0032] One or more processors;

[0033] A storage unit is used to store one or more programs, which, when executed by one or more processors, enable the one or more processors to implement the stacked board density identification method.

[0034] The beneficial effects of this invention are as follows:

[0035] 1. The width W and length L of the stacked boards are obtained by the distance sensor groups A and B set on the bracket assembly. The bracket assembly is a rectangular frame with scale lines. Two symmetrical distance sensors are set on each pair of opposite frames. The two distance sensors move along the frame by sliders to measure boards of different sizes.

[0036] 2. By setting up thickness detection components, weight detection components, and length and width detection components, the system automatically determines the length, width, and height dimensions of the stacked board blocks, the thickness of each board, the number of boards, and the weight of the stacked board blocks. This reduces tedious manual operations, saves manpower and resources, avoids errors caused by manual operations, lowers the human error rate, and improves automatic purification efficiency.

[0037] 3. At least two distance sensor groups are deployed on the width direction of the board stack block. By establishing a two-dimensional coordinate system, the coordinates of the four vertices of the board stack block in the preset coordinate system are obtained. The length L and width W of the board stack block are calculated through the coordinates. The height H is obtained through the distance sensor in the height direction. This is to solve the problem of large measurement error caused by the misalignment of the central axis of the board stack block and the bottom plate in the existing fully automated board cleaning process. The measurement images of two-sided or three-sided views of the board stack block are obtained through the camera component. The measurement images of the two-sided or three-sided views are preprocessed by cropping and correction (perspective transformation method) to obtain the side view of the board to be identified. Then, the thickness of the single board is identified by the convolutional neural network, which improves the accuracy of single board thickness identification in the existing fully automated board cleaning process. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the identification system of the present invention;

[0039] Figure 2 This is a schematic diagram of the material preparation platform of the present invention;

[0040] Figure 3 This is a schematic diagram of the first distance D0 of the present invention;

[0041] Figure 4 This is a schematic diagram of the first preset distance DX0 of the present invention;

[0042] Figure 5 This is a schematic diagram of the slide rail on the feeding side of the present invention;

[0043] Figure 6 This is a schematic diagram of the structure of multiple slide rails on the material preparation platform of the present invention;

[0044] Figure 7 This is a schematic diagram of the deployment of the length and width detection components of the present invention;

[0045] Figure 8 This is a schematic diagram of the two-dimensional coordinate system of the present invention;

[0046] Figure 9 This is a schematic diagram of the thickness calculation method of the present invention;

[0047] Figure 10 This is a schematic diagram of the margin calculation process of the present invention;

[0048] Figure 11 These are measured images of three views of the plate stacking block of the present invention;

[0049] Figure 12 This is a calibrated measurement image of the plate stacking block of the present invention;

[0050] Reference numerals: 1-Preparation platform, 2-Feeding platform, 3-Linear drive mechanism, 4-Sleeper, 5-Slot, 6-Blocking block, 7-Slide groove, 8-Slide rail, 9-Slider, 10-Distance sensor, 11-Mounting bracket, 12-Sheet stacking block, 13-Distance sensor group C, 14-Camera assembly. Detailed Implementation

[0051] 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. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. 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.

[0052] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention.

[0053] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0054] Furthermore, for clarity and brevity, descriptions of well-known structures, functions, and configurations may have been omitted. Those skilled in the art will recognize that various changes and modifications can be made to the examples described herein without departing from the spirit and scope of this disclosure.

[0055] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0056] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0057] Example 1

[0058] A stacked board density identification system includes a preparation platform for placing stacked boards and a thickness detection component, a weight detection component, and a length and width detection component, wherein:

[0059] The thickness detection component includes a distance sensor group C and a camera component disposed above the material preparation platform, wherein the distance sensor group C is used to measure the height of the stacked plates, and the camera is used to acquire images of the side of the stacked plate blocks;

[0060] The length and width detection component includes distance sensor groups A and B. Distance sensor group A includes distance sensors respectively installed on the feeding side and the side opposite to the feeding side on the top surface of the material preparation platform. Distance sensor group B includes two distance sensors respectively installed on the top surface of the material preparation platform perpendicular to the feeding side. It also includes a height distance sensor installed above the top surface of the material preparation platform and a controller.

[0061] The weight detection component includes a weight sensor installed on the bottom surface of the material preparation platform;

[0062] It also includes a controller, which is connected to the thickness detection component, the weight detection component, and the length and width detection component respectively, to receive data from the thickness detection component, the weight detection component, and the length and width detection component and perform density calculation.

[0063] Specifically, the distance sensor group C includes at least one distance sensor, the sensing direction of which is directed towards the top surface of the stack of boards, for detecting the distance between the sensor and the stack of boards in the height direction.

[0064] In one embodiment, the camera assembly is deployed diagonally above the stack of sheet metal blocks and includes a gimbal and a camera mounted on the gimbal, with the controller connected to the gimbal and the camera respectively.

[0065] Specifically, the camera and distance sensor group C are mounted on a mounting bracket, which is located on the diagonal extension line of the distance from the stacked plate blocks.

[0066] The camera can be connected to a controller to acquire measurement images of a predetermined area according to the controller's instructions. After the controller receives the image sent by the camera component, it starts to perform image recognition of the stacked blocks of boards in the image.

[0067] The controller is connected to the gimbal for controlling the gimbal. The gimbal is connected to the camera for controlling the camera's rotation. The MCU control circuit board is connected via a USB hub. The PC side mainly includes an interface layer and an application layer. The application layer is used to display the user UI interface, and the interface layer is used to implement face recognition and face detection algorithms. At the same time, the interface layer drives the control module through driver software.

[0068] Specifically, the camera assembly can remain on continuously, acquiring measurement images of a predetermined area at preset time intervals, such as acquiring images of the predetermined area for identification every hour. The camera assembly can also be off, but can be turned on upon receiving a drive command from the controller to acquire measurement images of the predetermined area. The camera assembly includes multiple cameras arranged vertically and installed at predetermined positions coinciding with the central axis of the stacked sheet material, allowing for image capture of the stacked sheet material at different heights.

[0069] Before the board material undergoes automated purification, the controller drives the distance sensor group C to measure the first preset distance DX0 between it and the material preparation platform.

[0070] When automated cleaning of the boards is required, the operator uses a forklift to transport the neatly stacked boards to the designated preparation platform. The controller sends detection commands to the distance sensor group and shooting commands to the camera component to obtain the first distance D0 measured by the distance sensor and to obtain multiple measurement images of the board stacks captured by the camera component.

[0071] In one embodiment of this application, when calculating the height H of the board stack block on the material preparation platform equipped with sleepers, it is also necessary to consider the thickness of the sleepers and the gap height between the sleepers and the material preparation platform. The thickness of the sleepers and the gap height between the sleepers and the material preparation platform can be obtained by actual measurement and set as Δh. Then, the height H of the board stack block is H = DX0 - D0 - Δh.

[0072] Specifically, the system also includes a feeding platform and a linear drive mechanism. There are two material preparation platforms. The linear drive mechanism is used to transport the corresponding material preparation platform to the feeding platform via the feeding side of the feeding platform.

[0073] Those skilled in the art will understand that the linear drive mechanism can also be, but is not limited to, a synchronous belt drive, a ball screw drive, a chain drive, or a tie rod drive, etc. The material preparation platform includes at least two sleepers mounted on its top, and the position of the sleepers on the platform perpendicular to the feeding side is adjustable. The number of sleepers can be freely selected according to the size of the material.

[0074] The feeding platform is rectangular, with three sides serving as the feeding sides, and a material preparation platform is provided on each of the three sides. A blocking block is also provided on the bottom side of the material preparation platform away from the feeding platform.

[0075] In this embodiment, the forklift places the stacked boards on the preparation platform. For ease of understanding, the preparation platforms are labeled 1, 2, ... etc. The linear drive mechanism moves the No. 1 preparation platform, which carries the stacked boards, to the feeding platform, where it waits for the artificial board processing unit to process the artificial boards. When the artificial boards on the No. 1 preparation platform have been processed, the linear drive mechanism 3 moves the No. 1 preparation platform back to its initial position from the feeding platform, and moves the No. 2 preparation platform, which carries the stacked boards, to the feeding platform, where it waits for the artificial board processing unit to process the artificial boards. This process continues until all the stacked materials on the preparation platforms 1, 2, ... have been processed, and a prompt is issued so that personnel can place the artificial boards again.

[0076] Specifically, the distance sensor group A includes four distance sensors A1, A2, A3, and A4. Distance sensors A1 and A2 are located on the feeding side, and distance sensors A3 and A4 are located on the side opposite to the feeding side. The sensing directions of distance sensors A1, A2, A3, and A4 during measurement are respectively facing the two parallel sides of the stacked plate block.

[0077] Specifically, the distance sensor group B includes four distance sensors B1, B2, B3, and B4. Each pair of distance sensors B1, B2, B3, and B4 is arranged in pairs. Each pair of distance sensors is respectively set on both sides of the top surface of the material preparation platform perpendicular to the feeding side, so that the sensing direction of the distance sensors A1 and A2 during measurement is directly facing the side of the stacked blocks of plates.

[0078] Preferably, each side of the top surface of the material preparation platform is provided with an embedded slide rail and two sliders disposed on the slide rail, and distance sensors A1, A2, A3, A4, B1, B2, B3, and B4 are respectively installed on the sliders.

[0079] When automated cleaning of sheet materials is required, operators use a forklift to transport neatly stacked sheet material blocks to a designated preparation platform, ensuring that the stacked blocks coincide with the centerline of the support assembly. By adjusting the sliders on the frame, the distance sensors on the frame are positioned at the stacked sheet material blocks. Scale lines ensure that the distance sensors on opposite frames are symmetrically positioned and measure symmetrically. The width of the stacked sheet material blocks is obtained through distance sensor group A, and the length is obtained through distance sensor group B.

[0080] Example 2

[0081] A method for identifying the density of stacked boards specifically includes the following steps:

[0082] Receive data from the thickness detection component and perform thickness calculation to obtain the thickness h of the single board and the number N of single boards in the board stack block;

[0083] Receive data from the length and width detection components and perform length and width calculations to obtain the length L and width W of the single board;

[0084] Receive weight data M from the weight detection component;

[0085] The density of the veneer is calculated based on the thickness h of the veneer, the number N of veneers in the stacked veneer block, the length L of the veneer, the width W of the veneer, and the weight M.

[0086] Specifically, the thickness h of the single board and the number N of single boards in the board stacking block include the following steps:

[0087] S001. Obtain the first distance D0 between the distance sensor group C and the top surface of the plate stack block, as measured by the distance sensor.

[0088] S002. Based on the first preset distance DX0 and the first distance D0 between the distance sensor group and the material preparation platform, calculate the height H of the stacked plate block as H = DX0 - D0.

[0089] S003. Acquire the measurement image captured by the camera, and determine whether the measurement image contains the sides of all the single boards;

[0090] When it is determined that the measurement image contains the sides of all the single boards, the number of boundary lines K in the measurement image is obtained by using a pre-established recognition model;

[0091] The number of individual boards N in the board stack block is calculated based on the number of boundary lines K.

[0092] When it is determined that the measured image does not include the sides of all the panels, a command is issued to adjust the camera's shooting direction.

[0093] Specifically, the number K of the dividing lines is the dividing line between two single boards, and the number of single boards in the board stack block N = K + 1.

[0094] When it is determined that the measured image does not include the sides of all the panels, a command is issued to adjust the camera's shooting direction.

[0095] S004. Based on the height H of the stacked board block and the number of individual boards N, calculate the thickness h of the individual board corresponding to the stacked board block, where h = H / N.

[0096] Specifically, the original coordinates of the four corner points are obtained through the following steps:

[0097] Line detection is performed in the measurement image to obtain the two endpoints of the line segments corresponding to each of the four sides in the measurement image;

[0098] Generate the equation of the line corresponding to each edge based on the two endpoints of the line segment corresponding to that edge;

[0099] Based on the equations of the lines corresponding to the four sides in the measured image, the coordinates of the four intersection points are obtained, and these four intersection point coordinates are determined as the original coordinates of the four corner points.

[0100] Specifically, the measurement image is a measurement image of the side of the stacked board blocks, acquired under the guidance of the placement area frame; the placement area frame is used to guide the camera to adjust the shooting angle so that the side area of ​​the stacked board blocks is located within the placement area frame for shooting;

[0101] The line detection in the measurement image includes: performing line detection in the measurement image within a predetermined range corresponding to each side of the placement region box.

[0102] Specifically, the acquisition of the measurement image containing the side of the stacked board blocks, guided by the placement area frame, includes the following steps:

[0103] Acquire the currently captured measurement image;

[0104] Determine whether the Intersection over Union (IOU) between the side area of ​​the stacked board block and the placement area frame in the currently captured measurement image is greater than a preset intersection threshold. If it is less than the threshold, prompt the user to adjust the camera shooting angle until the side area of ​​the stacked board block is located within the placement area frame.

[0105] Specifically, this device achieves 360° continuous rotation of the camera via a rotating gimbal. This includes:

[0106] Acquire a measurement image, and when the measurement image contains the side of a stack of boards, determine the first position of the side of the stack of boards in the measurement image;

[0107] Determine whether the first position meets the adjustment conditions. If so, determine the offset angle of the camera based on the second position and the first position. The second position includes the center position of the measured image, and the offset angle includes four directions: up, down, left, and right.

[0108] A control command is generated based on the offset angle, and the control command is used to control the pan-tilt unit connected to the camera to move the camera by the offset angle;

[0109] The adjustment condition is specifically the placement area frame comparison method: determine whether the IOU between the side area of ​​the stacked board block and the placement area frame in the currently captured measurement image is greater than the preset intersection threshold. If it is less than the threshold, prompt to adjust the camera shooting angle until the side area of ​​the stacked board block is located within the placement area frame.

[0110] The offset angle is the rotation angle of the camera, and the rotation direction can be up, down, left, or right, meaning the camera can rotate in four directions. When the camera offsets, the range of the captured image changes accordingly, and the position of the side of the stacked board block in the image changes accordingly. In one embodiment, the database stores the change distance of the first position corresponding to the offset angle in different directions. For example, if the camera rotates 5° in one direction (such as horizontally), the first position of the side of the stacked board block in the image will be translated by 2cm accordingly. Specifically, the distance value 'a' between the first position and the second position is calculated, and the x-axis and y-axis component distance values ​​'ax' and 'ay' of this distance value are calculated. Based on the relationship between the distance value and the offset angle stored in the database, the corresponding offset angle 'bx' for the 'ax' distance value and the corresponding offset angle 'by' for the 'ay' distance value are obtained, where the 'bx' offset angle is the angle of rotation offset along the x-axis and the 'by' offset angle is the angle of rotation offset along the y-axis. 'bx' and 'by' are determined as the offset angle that the camera needs to rotate at the current time.

[0111] For example, the control command is generated by the controller. If the determined offset angle bx along the x-axis is 5° and the offset angle by along the y-axis is 10°, then the control command is generated to control the gimbal. The camera on the gimbal is tilted, i.e., the control command causes the camera to rotate 5° and 10° along the x-axis and y-axis respectively. In another embodiment, the gimbal is a three-axis gimbal, and the overall rotation direction of the camera can be calculated. For example, it may rotate along a 30° angle between the x-axis and y-axis, meaning it does not rotate purely horizontally or vertically, but rather with a certain tilt angle.

[0112] This process, which begins before acquiring the measured images, also includes:

[0113] When a side recognition event is detected on the stacked board blocks, the camera is controlled to capture images.

[0114] If the measured image does not contain the side of the stacked plate block, then the measured image is used for part identification;

[0115] The first offset angle is determined based on the part recognition result, and a first control command is generated based on the first offset angle. The first control command is used to control the pan-tilt unit connected to the camera to move the camera at the first offset angle.

[0116] The step of determining a first offset angle based on the location identification result and generating a control command based on the first offset angle, wherein the control command is used to control the pan-tilt unit connected to the camera to move the camera at the first offset angle, includes:

[0117] Based on the part identification results, the corresponding offset angle range associated with the part is searched in the preset table;

[0118] A control instruction group is generated based on the offset angle range, and each control instruction in the control instruction group corresponds to an offset angle in the offset angle range.

[0119] The control command group is used to control the pan-tilt unit connected to the camera, causing the camera to move or rotate within the offset angle range.

[0120] The location identified is the area where the stacked plates are placed. Multiple identification locations can be placed around the material preparation platform in the automated stacking plate workshop, and different first offset angles can be determined according to different identification locations.

[0121] By efficiently tracking the stacked board blocks, the adjustment efficiency is higher, ensuring that the sides of the stacked board blocks are always within the placement area frame during each measurement process. This eliminates the need for manual adjustment by the user, reducing the difficulty of user operation and improving the accuracy of the shooting angle adjustment. This allows the detection of the original four corner coordinates to be within a predetermined range of each side of the placement area frame, enabling straight line detection in the measurement image and thus improving the speed of calculating the original four corner coordinates.

[0122] In this embodiment of the invention, when a stack of board materials in a preset area is identified by an image model, the controller inputs the stacked board materials in the preset area into a pre-established recognition model to obtain the number of boundary lines of the stacked board materials in the image of the preset area. This recognition model is stored in the controller and is activated after the measurement image is input. Furthermore, the recognition model can also be a deep learning model, including a deep learning unit and an output terminal. The recognition model uses the deep learning unit to receive and analyze the measurement image, and then outputs the analysis results from the output terminal. The deep learning unit can be an RNN learning unit, a CNN learning unit, or an LSTM learning unit, etc.

[0123] In this embodiment of the invention, the board stacking block is fixedly arranged within a predetermined area. For example, in automated board cleaning, the board stacking block is placed on a preparation platform by a forklift.

[0124] Those skilled in the art will understand that obtaining the number of boundary lines of the stacked board blocks in the measurement image by recognizing the model includes initially labeling the measurement images of the stacked board blocks for training, labeling the number of boundary lines in each measurement image, and the labeled image sample data includes a measurement image and a label for the number of boundary lines.

[0125] In one embodiment, a neural network is trained based on initially labeled image sample data;

[0126] Using a trained neural network, the confidence level of the number of stacked boards to be labeled at each boundary line is determined; the output of the neural network is the final estimate of the confidence level of the number of boards at each boundary line; based on these estimates, the number of boards with the highest confidence level is selected as the final output.

[0127] In one embodiment, a neural network is trained based on initially labeled image sample data;

[0128] Using a trained neural network, the confidence level of the number of stacked boards to be labeled at each boundary line is determined; the output of the neural network is the final estimate of the confidence level of the number of boards at each boundary line; based on these estimates, the number of boards with the highest confidence level is selected as the final output.

[0129] In one embodiment, the recognition model includes a side recognition model and a recognition boundary line model.

[0130] The measured image is input into the side recognition model, wherein the side recognition model is trained using multiple sets of training data, and each set of training data includes: an image of a stack of boards and identification information used to identify whether the side of the stack of boards is complete;

[0131] Obtain the output information of the model, wherein the output information includes an indicator of whether the measured image is complete.

[0132] Input the measured image into the plate stack block model;

[0133] The measurement image is obtained from the board stack block model to determine whether it contains information about all the sides of the veneer. The side recognition model is trained using multiple sets of data through machine learning. The multiple sets of data include a first type of data and a second type of data. Each set of data in the first type of data includes: an image containing information about all the sides of the veneer and a label indicating that the image contains information about all the sides of the veneer. Each set of data in the second type of data includes: a image not containing information about all the sides of the veneer and a label indicating that the image does not contain information about all the sides of the veneer.

[0134] When the measured image does not contain the stack of plates, a command is sent to the camera assembly to readjust the shooting angle.

[0135] Specifically, the length L and width W of the single board include the following steps:

[0136] S01. Acquire the data measured by distance haptic sensor group A and generate coordinate group AA; acquire the data measured by distance haptic sensor group B and generate coordinate group BB.

[0137] S02. Calculate the equations of the edge lines based on the coordinate set AA and coordinate set BB.

[0138] S03. Calculate the set of intersection points of the edge lines based on the set of edge line equations;

[0139] S04. Calculate the edge distance of the stacked board blocks based on the intersection point group, that is, the length L and width W of the single board.

[0140] Specifically, step S1 includes the following steps:

[0141] The distances between the sides corresponding to the stacked plate blocks are measured by the four distance sensors A1, A2, A3, and A4 of the distance sensor group A.

[0142] The distances between the sides of the stacked plate blocks are measured by the four distance sensors B1, B2, B3, and B4 of the distance sensor group B, respectively.

[0143] Specifically, the distances measured by distance sensors A1 and A2 are D1 and D2, and the distances measured by distance sensors A3 and A4 are D3 and D4.

[0144] The distances measured by distance sensors B1 and B2 are D5 and D6, and the distances measured by distance sensors B3 and B4 are D7 and D8.

[0145] Establish a two-dimensional coordinate system based on the material preparation platform;

[0146] The coordinates of each distance sensor in distance sensor group A and distance sensor group B in the two-dimensional coordinate system are obtained respectively, and the coordinates of the measurement point of the distance sensor on the side of the plate stack block are determined according to the distance measured by each distance sensor.

[0147] The coordinates of the measurement points of distance sensors A1, A2, A3, and A4 are defined as coordinate group AA;

[0148] The coordinates of the measurement points of distance sensors B1, B2, B3, and B4 are defined as coordinate group BB.

[0149] Specifically, step S02 includes the following steps:

[0150] The first sideline equation L1 is determined based on the coordinates of the measured points A1 and A2 in the coordinate set AA, and the second sideline equation L2 is determined based on the coordinates of the measured points A3 and A4 in the coordinate set AA.

[0151] The third sideline equation L3 is determined based on the coordinates of the measured points B1 and B2 in the coordinate set BB, and the fourth sideline equation L4 is determined based on the coordinates of the measured points B3 and B4 in the coordinate set BB.

[0152] The intersection point group is the intersection point group calculated based on the first sideline equation L1, the second sideline equation L2, the third sideline equation L3, and the fourth sideline equation L4, namely, four intersection points O1, O2, O3, and O4.

[0153] In one embodiment, a two-dimensional coordinate system is established based on the material preparation platform: a coordinate system is established with the X-axis as the X-axis and the side perpendicular to the feeding side as the Y-axis, so that the material preparation platform is located in the first quadrant of the two-dimensional coordinate system;

[0154] After each forklift places a stack of sheet metal, the distance between each distance sensor and the side of the corresponding base is measured to obtain the positional relationship of each distance sensor relative to the base, thereby determining the coordinates of each distance sensor in the two-dimensional coordinate system.

[0155] In one embodiment of the present invention, a camera is mounted on top of a stack of sheet metal blocks. When a stack of sheet metal blocks in a preset area is identified by an image model, the controller inputs the stack of sheet metal blocks in that preset area into a pre-established recognition model to obtain an image of the stack of sheet metal blocks in the image of that preset area. The preset correspondence between pixel values ​​and actual dimensions is stored in the controller's standard size database. The recognition model is stored in the controller and is activated after the input measurement image. Furthermore, the recognition model can also be a deep learning model, including a deep learning unit and an output terminal. The recognition model uses the deep learning unit to receive and analyze the measurement image, and then outputs the analysis results from the output terminal. The deep learning unit can be an RNN learning unit, a CNN learning unit, or an LSTM learning unit, etc. The above recognition model can be built using algorithms or application programs.

[0156] In this embodiment of the invention, the stacked board blocks are fixedly arranged within a predetermined area. For example, in automated board cleaning, the actual height value of the stacked board blocks is stored in the controller as a reference value for calculation.

[0157] Those skilled in the art will understand that by recognizing the model, the pixel values ​​of the stacked blocks of the board material in the measurement image are obtained, and each pixel value of the stacked blocks of the board material corresponds to a unique actual size value. By using a query matching method, the actual length and width values ​​of the stacked blocks of the board material in each frame of the measurement image are obtained.

[0158] In one embodiment, in the field of view from the camera to the board stack, a calibration window of the board stack is set at a position at a preset distance from the camera, wherein the calibration window has a predetermined physical window size; the physical board stack size of the board stack relative to the calibration window is estimated based on the window pixel value of the calibration window, the pixel value of the board stack, and the physical window size, wherein the pixel ratio and physical size ratio of the window size to the board stack size are equal.

[0159] In addition, stickers or other markers with known dimensions can be affixed to the preset area as a reference; this is not limited to this method. Furthermore, the calibration window is movably positioned within the predetermined area, meaning it is movable and can be placed at any location within the predetermined area by the operator when measuring the height of stacked sheet materials. For example, the reference object can be a cube with known dimensions stored in the controller.

[0160] In one embodiment, the pixel height of the board stack is obtained in each frame of the measurement image by capturing measurement images at different heights of the board stack block using multiple cameras. A search method is then used to match a height Hi that matches each pixel height, and a correction factor height for the board stack block is calculated. The height value H measured by the height distance sensor is corrected using the height H0 of the plate stack block to obtain the height H of the plate stack block.

[0161] Example 3

[0162] A control device for identifying the density of stacked boards, characterized in that it comprises:

[0163] One or more processors;

[0164] A storage unit is used to store one or more programs, which, when executed by one or more processors, enable the one or more processors to implement the stacked board density identification method.

[0165] Example 4: A computer-readable storage medium storing a computer program thereon, which, when executed by a processor, enables the implementation of the aforementioned method for identifying the density of stacked materials.

[0166] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Based on the technical essence of the present invention, any simple modifications, equivalent substitutions, and improvements made to the above embodiments within the spirit and principles of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A system for identifying the density of stacked boards, characterized in that, This includes a material preparation platform for placing stacked boards, as well as thickness detection components, weight detection components, and length and width detection components. A two-dimensional coordinate system is established based on the material preparation platform, with the X-axis being the side edge of the feeding side and the Y-axis being the side edge perpendicular to the feeding side. The thickness detection component includes a distance sensor group C and a camera assembly disposed above the material preparation platform, wherein the distance sensor group C is used to measure the height of the stacked boards, and the camera is used to acquire images of the side of the stacked board blocks; it also includes a height and distance sensor disposed above the top surface of the material preparation platform and a controller; The camera assembly includes multiple cameras, which are deployed diagonally above the stack of boards along the height direction, and each camera forms a different angle with the center of the top surface of the stacked boards, for capturing measurement images of the side of the stack of boards from different heights. The length and width detection component includes distance sensor groups A and B. Distance sensor group A includes distance sensors respectively installed on the feeding side and the side opposite to the feeding side on the top surface of the material preparation platform. Distance sensor group B includes two distance sensors respectively installed on the top surface of the material preparation platform perpendicular to the feeding side. The distance sensor group A includes four distance sensors A1, A2, A3, and A4. Distance sensors A1 and A2 are located on the feeding side, and distance sensors A3 and A4 are located on the side opposite to the feeding side. The sensing directions of distance sensors A1, A2, A3, and A4 during measurement are respectively facing the two parallel sides of the stacked plate block. The distance sensor group B includes two distance sensors respectively set on the top surface of the material preparation platform perpendicular to the feeding side. The distance sensor group B includes four distance sensors B1, B2, B3, and B4. The distance sensors B1, B2, B3, and B4 are paired up in pairs. Each pair of distance sensors is set on the top surface of the material preparation platform perpendicular to the feeding side, so that the sensing direction of the distance sensors A1 and A2 during measurement is directly facing the side of the stacked plate block. The weight detection component includes a weight sensor installed on the bottom surface of the material preparation platform; It also includes a controller, which is connected to the thickness detection component, the weight detection component, and the length and width detection component, respectively. The controller performs the following steps: The thickness is calculated by receiving data from the thickness detection component to obtain the thickness h of the single board and the number N of single boards in the board stack, where h = H / N, and H represents the height of the board stack. The calculation process is as follows: Measurement images are captured by multiple cameras at different heights of the board stack. The pixel height of the board stack in each frame of the measurement image is obtained. A search method is used to match the height Hi that matches each pixel height, and the board stack correction factor height is calculated. The height value H measured by the height distance sensor is corrected using the plate stacking block correction factor height H0, thereby obtaining the height H of the plate stacking block; Receive weight data M from the weight detection component; The process of receiving data from the length and width detection component and calculating the length and width to obtain the board's length L and width W is as follows: Receive data from the distance sensor group A1, A2, A3, A4, B1, B2, B3, B4 to calculate the edge line equations of the stacked board blocks, calculate the intersection point group of the edge lines based on the edge line equations, and calculate the edge distance of the stacked board blocks based on the intersection point group, i.e., the length L and width W of a single board. Specifically, data measured by distance sensor group A is acquired and coordinate group AA is generated, and data measured by distance sensor group B is acquired and coordinate group BB is generated; Calculate the equations of the edge lines based on the coordinate set AA and coordinate set BB; The distances between the sides corresponding to the stacked plate blocks are measured by the four distance sensors A1, A2, A3, and A4 of the distance sensor group A. The distances between the sides of the stacked plate blocks are measured by the four distance sensors B1, B2, B3, and B4 of the distance sensor group B, respectively. The coordinates of each distance sensor in distance sensor group A and distance sensor group B in the two-dimensional coordinate system are obtained respectively, and the coordinates of the measurement point of the distance sensor on the side of the plate stack block are determined according to the distance measured by each distance sensor. The coordinates of the measurement points of distance sensors A1, A2, A3, and A4 are defined as coordinate group AA; The coordinates of the measurement points of distance sensors B1, B2, B3, and B4 are defined as coordinate group BB; Calculate the density ρ of the veneer.

2. The stacked board density identification system according to claim 1, characterized in that, The distance sensor group C includes at least one distance sensor, the sensing direction of which is directed towards the top surface of the stack of boards, for detecting the distance between the sensor and the stack of boards in the height direction.

3. The stacked board density identification system according to claim 1, characterized in that, The camera assembly is deployed diagonally above the stack of plates and includes a gimbal and a camera mounted on the gimbal. The controller is connected to the gimbal and the camera respectively.

4. The stacked board density identification system according to claim 2, characterized in that, Each side of the top surface of the material preparation platform is provided with an embedded slide rail and two sliders on the slide rail. Distance sensors A1, A2, A3, A4, B1, B2, B3, and B4 are respectively installed on the sliders.

5. A method for identifying the density of stacked boards, said identification method being based on a stacked board density identification system according to any one of claims 1-4, characterized in that, Specifically, the following steps are included: Receive data from the thickness detection component and perform thickness calculation to obtain the thickness h of the single board and the number N of single boards in the board stack block; Receive data from the length and width detection components and perform length and width calculations to obtain the length L and width W of the single board; Receive weight data M from the weight detection component; The density of the veneer is calculated based on the thickness h of the veneer, the number N of veneers in the stacked veneer block, the length L of the veneer, the width W of the veneer, and the weight M.

6. The method for identifying the density of stacked plates according to claim 5, characterized in that, The thickness h of the single board and the number N of single boards in the board stacking block specifically include the following steps: S001. Obtain the first distance D0 between the distance sensor in the distance sensor group C and the top surface of the plate stack block; S002. Based on the first preset distance DX0 and the first distance D0 between the distance sensor group and the material preparation platform, calculate the height H of the stacked plate block as H = DX1 - D0. S003. Acquire the measurement image captured by the camera, and determine whether the measurement image contains the sides of all the single boards; When it is determined that the measurement image contains the sides of all the single boards, the number of boundary lines K in the measurement image is obtained by using a pre-established recognition model; The number of individual boards N in the board stack block is calculated based on the number of boundary lines K. S004. Based on the height H of the stacked board block and the number of individual boards N, calculate the thickness h of the individual board corresponding to the stacked board block, where h = H / N.

7. The method for identifying the density of stacked plates according to claim 5, characterized in that, The length L and width W of the single board are specifically defined by the following steps: S01. Acquire the data measured by distance haptic sensor group A and generate coordinate group AA; acquire the data measured by distance haptic sensor group B and generate coordinate group BB. S02. Calculate the equations of the edge lines based on the coordinate set AA and coordinate set BB. S03. Calculate the set of intersection points of the edge lines based on the set of edge line equations; S04. Calculate the edge distance of the stacked board blocks based on the intersection point group, that is, the length L and width W of the single board.

8. A control device for identifying the density of stacked boards, characterized in that, include: One or more processors; A storage unit for storing one or more programs, which, when executed by one or more processors, enable the one or more processors to implement a method for identifying the density of stacked boards according to any one of claims 5 to 7.

Images