An artificial intelligence wood chip automatic grading method and system based on deep learning

Abstract

The present invention discloses a method and system for automatic grading of wood chips using artificial intelligence based on deep learning, wherein the system comprises: an equipment frame, a loading component, a quality grading detection module, a unloading component, and a main control system module; the equipment frame is used to carry the components in the grading system; the loading component is used to grab the wood chips to be graded and place them in the main control system module for screening and guiding; the guide plate sorting mechanism is used to sort the wood chips by grade, and the main control system module is used to control the grading system to achieve automatic grading. The above-mentioned system and method can realize automatic and efficient identification of wood chip surface defects and automatic grading and screening of wood chips, avoiding the problems of large errors, low efficiency, and high cost of manual grading, making wood chip sorting automated and large-scale, and having broad market application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of wood processing and application development, and in particular to an artificial intelligence wood chip automatic grading method and system based on deep learning. Background Art

[0002] Plywood is one of the most important wood products. Its production process involves peeling, peeling, drying, gluing, and molding logs to create plywood. However, during the wood's growth cycle, defects such as knots and wormholes are easily formed. Furthermore, during the drying process, peeled wood chips are affected by factors such as wood growth stress and water evaporation rate, resulting in surface shrinkage and cracking. These factors not only affect the plywood's appearance but also severely reduce its strength, ultimately affecting its quality.

[0003] During the wood processing process, it is crucial to classify wood chips into different grades. Single sheets with fewer surface defects are used for plywood panels, while those with more defects are used for core boards or lower-end plywood products. Therefore, surface defects in wood chips are a key indicator of wood quality. As the wood processing industry evolves towards mechanized and automated large-scale production, companies are placing increasing emphasis on the quality of board processing, particularly surface defects. Consequently, surface defect detection technology for wood chips has become increasingly important. However, many small processing companies rely on manual grading for chip grading. This method relies on visual inspection and work experience to determine the type of surface defects and grade the veneer. This method, however, is subject to significant subjective influences based on individual experience, and different staff members may have different judgments. This not only reduces work efficiency but also affects product grading results due to the influence of individual skills. Misjudgments in chip grading can lead to losses in subsequent processing.

[0004] Therefore, it is necessary to establish an efficient and reliable wood chip surface defect grading system based on artificial intelligence and machine vision to improve the product quality of plywood and reduce production costs. Summary of the Invention

[0005] In view of this, the present invention discloses an artificial intelligence wood chip automatic grading method and system based on deep learning, which is used to improve the product quality of plywood and reduce production costs.

[0006] In the first aspect, the present invention provides an artificial intelligence wood chip automatic grading system based on deep learning, specifically: including: an equipment frame, a loading component, a quality grading detection module, a unloading component, and a main control system module; the equipment frame is used to carry the components in the grading system; the loading component is used to grab the wood chips to be graded and place them in the main control system module for screening and guiding; the guide plate sorting mechanism is used to sort the wood chips according to grade, and the main control system module is used to control the grading system to achieve automatic grading.

[0007] Furthermore, the loading component includes a loading frame, a six-degree-of-freedom robotic arm, a pneumatic suction cup array, and a distance sensor;

[0008] The quality grading inspection module includes a 3D camera and a testing room;

[0009] The blanking components include a guide plate sorting mechanism and a wood chip collecting frame;

[0010] The main control system module includes an industrial computer, a first servo motor, and a first conveyor belt;

[0011] The equipment frame includes three parts from front to back: a first conveying part, a guide part, and a second conveying part;

[0012] The six-degree-of-freedom robotic arm is arranged on one side of the front end of the first conveying part, the distance sensor is installed at the end of the six-degree-of-freedom robotic arm, and the pneumatic suction cup array is installed at the clamping end of the six-degree-of-freedom robotic arm; the control motor and distance sensor of the six-degree-of-freedom robotic arm are respectively connected to the industrial computer signal.

[0013] The inspection room is arranged in the middle and rear part of the first conveyor, with both ends fixed to the equipment frame, and the 3D camera is fixed on the top of the inspection room; the 3D camera is connected to the industrial computer signal;

[0014] The first conveying part and the second conveying part are both provided with a first servo motor at the head and tail ends, and the first conveying belts in the first conveying part and the second conveying part are driven by the first servo motor to move;

[0015] The guide portion is a guide plate sorting mechanism, which includes a guide plate frame, a second conveyor belt, a second servo motor, and a third servo motor. The second conveyor belt is sleeved on the guide plate frame, the motor shaft of the second servo motor is connected to the second conveyor belt drive shaft, and the third servo motor is fixed to the equipment frame, and the motor shaft of the third servo motor is connected to the middle part of the guide plate frame. The wood chip collection frame is arranged at a corresponding position below the guide plate sorting mechanism.

[0016] The first servo motor, the second servo motor and the third servo motor are respectively connected to industrial computer signals.

[0017] Furthermore, the guide plate sorting mechanism is provided with three, namely, the A-level guide plate sorting mechanism, the B-level guide plate sorting mechanism, and the C-level guide plate sorting mechanism; correspondingly, the A-level collection frame, the B-level collection frame, and the C-level collection frame are provided under the guide plate sorting mechanism.

[0018] In a second aspect, the present invention provides a method for automatically grading wood chips using the above system, comprising the following steps:

[0019] S1: Identify surface defects of wood chips and grade them;

[0020] S2: Accurately locate the position of the wood chips:

[0021] S3: conveying and controlling wood chips, and finally sorting the graded wood chips;

[0022] Wherein said S1 comprises:

[0023] S11: Acquire a black and white image of the wood chip;

[0024] S12: Obtaining defect outer contour points of the black and white image according to the black and white image;

[0025] S13: Acquire the defect contour of the image;

[0026] S14: Calculate the defect area of ​​the image;

[0027] S15: Determine the grade of the wood chips.

[0028] Furthermore, S11 specifically includes: the information of the wood chip image actually collected includes: color data of red, green and blue, which is described as a five-dimensional array as Color_Image[i,j,r,g,b], where i represents the horizontal sequence number of the current pixel; j represents the vertical sequence number of the current pixel; r represents the red value of the pixel; g represents the green value of the pixel; and b represents the blue value of the pixel;

[0029] Color screen the image information of the wood chips, complete the setting of multiple pixel points, and obtain a black and white image, which is composed of pixel points in both horizontal and vertical directions: set the critical values ​​of the red, green and blue colors of the image (Cr, Cg, Cb), and judge the red, green and blue colors of all pixel points in the image: if the red, green and blue color values ​​of the pixel point are all higher than the critical values ​​of the red, green and blue colors (Cr, Cg, Cb), then the red, green and blue pixel values ​​of the point are all set to 0, otherwise the red, green and blue pixel values ​​of the point are all set to 1.

[0030] Furthermore, S12 is: if the RGB values ​​of Color_Image[i-1,j,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i+1,j,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i,j-1,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i,j+1,r,g,b] are all equal to 1, then the point is judged to be an internal point of the defect area, and the RGB values ​​of the point are set to 1, otherwise they are set to 0.

[0031] Furthermore, the S13 is: take the first pixel point N[i,j] whose rgb value is equal to 0 in the image array Color_Image as the current point, where i is the horizontal serial number of the pixel point and j is the vertical serial number of the pixel point; and record N[i,j] points in the outline array Outline[l,k,i,j], and judge the value of the pixel point in the order of pixel points N[i,j-1], N[i-1,j-1], N[i-1,j], N[i-1,j+1], N[i,j+1], N[i+1,j+1], N[i+1,j], N[i-1,j+1], where l is the serial number of the current outline, its initial value is 0, and its value increases by 1 for each additional outline; k is the lth The current serial number of the contour point, its initial value is 0, each time a point is recorded, the k value automatically increases by 1; i is the horizontal serial number of the pixel point, j is the vertical serial number of the pixel point, if the rgb values ​​of Color_Image[x,y,r,g,b] are all equal to 0, where x is the horizontal serial number of the above 8 pixel points; y is the vertical serial number of the above 8 pixel points, then record N[x,y] points in the contour array Outline[l,k,i,j], and set the rgb value of Color_Image[x,y,r,g,b] at this point to 1; repeat the above steps until all rgb values ​​in the Color_Image array are equal to 1, and obtain the Outline[l,k,i,j] contour array.

[0032] Furthermore, the S14 is as follows: since the outline of the defect on the surface of the wood chip is a convex polygon, the number of all points of Outline[l,k,i,j] is k, and k-1 triangles can be formed by connecting lines from the point Outline[l,1,i,j] to points other than the point Outline[l,2,i,j] and the end point; for each triangle, the lengths of the three sides a, b, c and the average perimeter p = (a+b+c) / 2 are calculated, according to the formula: the area of ​​the triangle s = 1 / 2 of p(pa)(pb)(pc); the areas of all k-1 triangles are counted to obtain the area S of the defect, and the number of defects Num = Num+1 and the sum of the defect areas Ssum = Ssum+S are counted;

[0033] After completing steps S11-S14 above, find the point where the pixel value v is not zero in the three-dimensional array Color_Image again, and execute steps S11-14 until all the pixels in the Color_Image array are 1; finally, the number Num of all defects on the wood chip and the sum of all defect areas S are obtained. sum .

[0034] Furthermore, the step S15 is: preset the critical defect number N of the grade A plate a , B-grade board critical defect number N b and critical defect area S of grade A plate a , B-grade plate critical defect area (S b );

[0035] When the number of defects on the wood chip Num <N a And the defect area S sum a It is judged as Grade A board; when the number of defects on the wood chip is N a <Num<N b And the defect area S a sum b It is judged as Grade B board; when the number of defects on the wood piece Num>N b And the defect area S sum >S b It is judged to be a C-grade board.

[0036] Furthermore, the method of S2 is as follows: measuring the distance between the pneumatic suction cup array at the end of the robotic arm and the wood chip using a distance sensor, wherein the distance sensor measures the propagation time t of the emitted sound wave or light wave between the pneumatic suction cup and the wood chip, and multiplies the propagation time t by the speed of the sound wave or light wave to obtain the distance between the wood chip and the pneumatic suction cup;

[0037] The method of S3 is as follows: a first servo motor drives a first conveyor belt to move the wood chips. The wood chips pass through a detection chamber and a guide plate sorting mechanism in sequence during the conveyor belt conveyance. After the industrial computer determines the grade of the wood chips, it sends a motion instruction to the corresponding guide plate sorting mechanism to rotate and guide the wood chips to fall into the corresponding wood chip collection frame, thereby completing the sorting operation. The initial position of the guide plate sorting mechanism is horizontal.

[0038] ​​​The present invention provides an artificial intelligence-based deep learning-based automatic wood chip grading method and system, which is a control method that includes efficient identification of wood chip surface defects, precise positioning of wood chips, and transmission and sorting of wood chips. Under the coordinated action of four parts: a loading component, a quality grading detection module, a unloading component, and a main control system module, the method realizes automatic and efficient identification of wood chip surface defects and automatic grading and screening of wood chips, avoiding the problems of large errors, low efficiency, and high cost of manual grading, making wood chip sorting automated and large-scale, and has broad market application prospects.

[0039] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0042] Figure 1 A schematic diagram of the composition structure of an artificial intelligence wood chip automatic grading system based on deep learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of systems consistent with certain aspects of the present invention, as detailed in the appended claims.

[0044] To address the issues of low efficiency and poor precision in manual sorting of wood chip surface defects, this implementation scheme provides an artificial intelligence wood chip automatic grading system based on deep learning, comprising: an equipment frame, a loading component, a quality grading detection module, a unloading component, and a main control system module; the equipment frame is used to carry the components in the grading system; the loading component is used to grab the wood chips to be graded and place them in the main control system module for screening and guiding; the guide plate sorting mechanism is used to sort the wood chips by grade, and the main control system module is used to control the grading system to achieve automatic grading.

[0045] like Figure 1As shown, the loading components include a loading frame 1, a six-degree-of-freedom robotic arm 2, a pneumatic suction cup array 3, and a distance sensor 4;

[0046] The above-mentioned quality grading inspection module includes a 3D camera 5 and an inspection chamber 6;

[0047] The above-mentioned unloading components include a guide plate sorting mechanism 7 and a wood chip collecting frame 8;

[0048] The main control system module includes an industrial computer 9, a first servo motor 10, and a first conveyor belt 11;

[0049] The above-mentioned equipment frame includes three parts from front to back: a first conveying part, a guide part, and a second conveying part;

[0050] The six-degree-of-freedom robotic arm 2 is arranged on one side of the front end of the first conveying section, the distance sensor 4 is installed at the end of the six-degree-of-freedom robotic arm 2, and the pneumatic suction cup array 3 is installed at the clamping end of the six-degree-of-freedom robotic arm 2; the control motor of the six-degree-of-freedom robotic arm 2, the distance sensor 4, and the pneumatic suction cup array 3 are respectively connected to the industrial computer 9 for signal communication;

[0051] The six-DOF robotic arm 2 is driven by six drive motors, moving the pneumatic suction cup array 3 above the wood chip 1. A distance sensor 4 measures the distance between the pneumatic suction cup array 3 and the wood chip and transmits this distance to the industrial computer 9. The industrial computer 9 sends motion control instructions to the six-DOF robotic arm 2, bringing the suction cup array close to the wood chip and controlling the vacuum solenoid valve to close. The suction cups generate suction, thereby grabbing the wood chip. The six drive motors then coordinate the movement of the six-DOF robotic arm 2, placing the wood chip 1 onto the conveyor belt 11.

[0052] The inspection room 6 is arranged in the middle and rear part of the first conveyor, with its two ends fixed to the equipment frame, and the 3D camera 5 is fixed on the top of the inspection room 6; the 3D camera 5 is connected to the industrial computer 9 by signal;

[0053] A first servo motor 10 is provided at both the head and tail ends of the first conveying part and the second conveying part, and the first servo motor drives the first conveyor belt 11 in the first conveying part and the second conveying part to move;

[0054] The guide part is a guide plate sorting mechanism 7, which includes two servo motors. One servo motor is inside the guide plate sorting mechanism, and the motor shaft is connected to the synchronous belt drive shaft, which is used to drive the conveyor belt of the guide plate sorting mechanism to rotate; the other servo motor is fixed on the equipment frame, and the motor shaft is connected to the guide plate sorting mechanism. Through the rotation of the servo motor, the guide plate sorting mechanism 7 can be driven to rotate clockwise and counterclockwise.

[0055] Specifically, the guide plate sorting mechanism 7 includes a guide plate frame, a second conveyor belt 12, a second servo motor 13, and a third servo motor 14. The second conveyor belt 12 is sleeved on the guide plate frame, the motor shaft of the second servo motor 13 is connected to the drive shaft of the second conveyor belt 12, the third servo motor 14 is fixed to the equipment frame, and the motor shaft of the third servo motor 14 is connected to the middle part of the guide plate frame; the wood chip collection frame 8 is arranged at a corresponding position below the guide plate sorting mechanism 7;

[0056] The first servo motor 10 , the second servo motor 13 , and the third servo motor 14 are all connected to the industrial computer 9 for signal signals.

[0057] Further improvement, the guide plate sorting mechanism 7 is provided with three, namely, the A-level guide plate sorting mechanism, the B-level guide plate sorting mechanism, and the C-level guide plate sorting mechanism; correspondingly, the A-level collection frame, the B-level collection frame, and the C-level collection frame are provided under the guide plate sorting mechanism respectively;

[0058] On the other hand, this embodiment adopts the automatic chip grading method of the above system, which realizes automatic chip grading by synergistically adopting the steps of efficiently identifying surface defects of wood chips, accurately locating the position of wood chips, and conveying and sorting wood chips.

[0059] Based on the above system structure, when sorting wood chips, the six-degree-of-freedom robot arm 2 moves to the top of the loading frame 1, and then the distance sensor 4 placed at the end of the six-degree-of-freedom robot arm 2 measures the distance between the end of the robot arm and the loading frame 1, and transmits the distance data to the industrial computer 9; the industrial computer 9 sends the position coordinates of the wood chips in the loading frame 1 to the six-degree-of-freedom robot arm 2, and the six-degree-of-freedom robot arm 2 drives the pneumatic suction cup array 3 to move above the wood chips, picks up the wood chips, and places the wood chips on the conveyor belt 11; then the industrial computer 9 controls the servo motor 10 to drive the conveyor belt 11 to send the wood chips into the inspection chamber 6; the camera 5 placed on the top of the inspection chamber 6 collects image information of the wood chips and transmits the image information data to the industrial computer 9. The industrial computer 9 counts the number of knots and cracks on the surface of the wood chips according to the following method, and after analysis, evaluates the grade of the wood chips, and controls the guide plate sorting mechanism to place wood chips of different grades into wood chip collection frames of different grades.

[0060] The specific steps include:

[0061] S1: Identify surface defects of wood chips and grade them;

[0062] S2: Accurately locate the position of the wood chips:

[0063] S3: conveying and controlling wood chips, and finally sorting the graded wood chips;

[0064] S1 includes:

[0065] S11: Acquire a black and white image of the wood chip;

[0066] The image data obtained by the industrial computer 9 is color data containing red, green and blue, and can be described by a five-dimensional array, Color_Image[i,j,r,g,b], where i represents the horizontal serial number of the current pixel; j represents the vertical serial number of the current pixel; r represents the red value of the pixel; g represents the green value of the pixel; and b represents the blue value of the pixel.

[0067] The collected wood chip surface image is subjected to color screening. The specific method is to set the critical values ​​of the red, green and blue colors of the image (Cr, Cg, Cb), and judge the red, green and blue colors of all pixels in the image. If the red, green and blue color values ​​of the pixel point are all higher than the critical values ​​of the red, green and blue colors (Cr, Cg, Cb), then the red, green and blue pixel values ​​of the point are set to 0. Otherwise, the red, green and blue pixel values ​​of the point are set to 1. After completing the setting of all pixel points, a black and white image is obtained, which is composed of pixels in both the horizontal and vertical directions.

[0068] S12: Obtaining the defect outer contour points of the black and white image based on the black and white image;

[0069] Analyze each pixel of the black and white image obtained in S11. If the RGB values ​​of Color_Image[i-1,j,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i+1,j,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i,j-1,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i,j+1,r,g,b] are all equal to 1, then the point is judged to be an internal point of the defect area and the RGB values ​​of the point are set to 1. Otherwise, they are set to 0. Analyze all pixel points to obtain the points of the outer contour of the defect.

[0070] S13: Acquire the defect contour of the image;

[0071] Take the first pixel point N[i,j] (i is the horizontal serial number of the pixel point; j is the vertical serial number of the pixel point) in the image array Color_Image with both rgb values ​​equal to 0 as the current point, and record N[i,j] points in the outline array Outline[l,k,i,j] (l is the serial number of the current outline, its initial value is 0, and its value increases by 1 for each additional outline; k is the current serial number of the lth outline point, its initial value is 0, and the k value automatically increases by 1 for each point recorded; i is the horizontal serial number of the pixel point; j is the vertical serial number of the pixel point), and record the points N[i,j] according to the pixel points N[i,j-1], N[i-1,j- 1],N[i-1,j],N[i-1,j+1],N[i,j+1],N[i+1,j+1],N[i+1,j],N[i-1,j+1],in order to judge the value of the pixel. If the rgb values ​​of Color_Image[x,y,r,g,b] (x is the horizontal serial number of the above 8 pixels; y is the vertical serial number of the above 8 pixels) are all equal to 0, then record N[x,y] points in the outline array Outline[l,k,i,j] and set the rgb values ​​of Color_Image[x,y,r,g,b] at this point to 1. Repeat the above steps until all rgb values ​​in the Color_Image array are equal to 1, and get the outline array Outline[l,k,i,j].

[0072] S14: Calculate the defect area of ​​the image;

[0073] Since the outlines of defects on the surface of wood chips are mostly convex polygons, we can form k-1 triangles by connecting all the points of Outline[l,k,i,j] (a total of k points) from Outline[l,1,i,j] to the points other than Outline[l,2,i,j] and the end point. For each triangle, calculate the lengths of the three sides: a, b, c, and the average perimeter p = (a+b+c) / 2, according to the formula: triangle area s = p(pa)(pb)(pc) to the power of 1 / 2. Count the areas of all k-1 triangles to obtain the area S of the defect, and count the number of defects Num = Num+1 and the sum of the defect areas S. sum =S sum +S.

[0074] After completing steps S11-S14 above, find the point where the pixel value v is not zero in the three-dimensional array Color_Image again, and execute steps 1-4 until all the pixels in the Color_Image array are 1. Finally, the number of all defects Num on the wood chip and the sum of all defect areas S are obtained. sum .

[0075] S15: Determine the grade of the wood chips.

[0076] Preset critical defect number of grade A plate (N a ), critical defect number of grade B plate (N b ) and critical defect area of ​​grade A plate (S a ), critical defect area of ​​Class B plate (S b ), when the number of defects on the wood chip Num <N a And the defect area S sum a It is judged as Grade A board; when the number of defects on the wood chip is N a <Num<N b And the defect area S a sum b It is judged as Grade B board; when the number of defects on the wood piece Num>N b And the defect area S sum >S b It is judged to be a C-grade board.

[0077] S2: Accurately locate the position of the wood chips:

[0078] The distance between the pneumatic suction cup array 3 at the end of the robotic arm and the wood chip is measured using a distance sensor 4. The distance sensor uses an ultrasonic sensor or a laser sensor. The working principle is to measure the propagation time (t) of the emitted sound or light wave between the pneumatic suction cup and the wood chip, and multiply the propagation time t by the speed of the sound wave or light wave to obtain the distance between the wood chip and the pneumatic suction cup.

[0079] S3: conveying and controlling wood chips, and finally sorting the graded wood chips;

[0080] ​​​There are three guide plate sorting mechanisms 7 on the sorting conveyor belt, corresponding to grade A, grade B, and grade C boards respectively. The initial positions of the three guide plates are all horizontal. The conveyor belt driven by the servo motor 10 drives the wood chips to move. During the movement, the wood chips pass through the detection room 6, the guide plate sorting mechanism 7 and other areas in sequence. After the industrial computer 9 determines the grade of the wood chips, it sends a movement instruction to the corresponding guide plate sorting mechanism 7. If it is grade A board, the No. 1 guide plate sorting mechanism 7 rotates -45 degrees clockwise, and after the wood chips are transported, it rotates 10-90 degrees counterclockwise and returns to the initial position; if it is grade A board, the No. 1 guide plate sorting mechanism 7 rotates clockwise The guide plate sorting mechanism 7 rotates -10-(-90) degrees, and after the wood chips are conveyed, it rotates 45 degrees counterclockwise to return to its initial position. For Class B boards, the second guide plate sorting mechanism 7 rotates -10-(-90) degrees clockwise, and after the wood chips are conveyed, it rotates 10-(-90) degrees counterclockwise to return to its initial position. For Class C boards, the third guide plate sorting mechanism 7 rotates -10-(-90) degrees clockwise, and after the wood chips are conveyed, it rotates -10-(-90) degrees counterclockwise to return to its initial position. The rotation of the guide plate sorting mechanism 7 guides the wood chips into the corresponding wood chip collection frame 8, completing the sorting operation.

[0081] Other embodiments of the present invention will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the claims.

Claims

1. An artificial intelligence wood chip automatic grading method based on deep learning, characterized in that: The deep learning-based AI-powered automatic wood chip grading equipment includes: an equipment frame, a loading component, a quality grading detection module, a unloading component, and a main control system module. The equipment frame is used to support the components of the grading system. The loading component is used to grab the wood chips to be graded and place them in the main control system module for screening and guidance. The guide plate sorting mechanism is used to sort the wood chips by grade, and the main control system module is used to control the grading system to achieve automatic grading. The loading component comprises a loading frame (1), a six-degree-of-freedom robotic arm (2), a pneumatic suction cup array (3), and a distance sensor (4); The quality grading inspection module includes a 3D camera (5) and a detection chamber (6); The unloading component includes a guide plate sorting mechanism (7) and a wood chip collecting frame (8); The main control system module includes an industrial computer (9), a first servo motor (10), and a first conveyor belt (11); The equipment frame comprises three parts from front to back: a first conveying part, a guide plate sorting mechanism (7), and a second conveying part; The six-degree-of-freedom robotic arm (2) is arranged on one side of the front end of the first conveying part, the distance sensor (4) is installed at the end of the six-degree-of-freedom robotic arm (2), and the pneumatic suction cup array (3) is installed at the clamping end of the six-degree-of-freedom robotic arm (2); the control motor of the six-degree-of-freedom robotic arm (2) and the distance sensor (4) are respectively connected to the industrial computer (9) for signal transmission; The inspection room (6) is arranged in the middle and rear part of the first conveying part, and its two ends are fixed on the equipment frame. The 3D camera (5) is fixed on the top of the inspection room (6); the 3D camera (5) is connected to the industrial computer (9) by signal. A first servo motor (10) is provided at both the head and tail ends of the first conveying part and the second conveying part, and the first servo motor drives the first conveyor belt (11) in the first conveying part and the second conveying part to move; The guide plate sorting mechanism (7) comprises a guide plate frame, a second conveyor belt (12), a second servo motor (13), and a third servo motor (14); the second conveyor belt (12) is sleeved on the guide plate frame; the motor shaft of the second servo motor (13) is connected to the drive shaft of the second conveyor belt (12); the third servo motor (14) is fixed to the equipment frame; the motor shaft of the third servo motor (14) is connected to the middle of the guide plate frame; the wood chip collecting frame (8) is arranged at a corresponding position below the guide plate sorting mechanism (7); The first servo motor (10), the second servo motor (13), and the third servo motor (14) are all respectively connected to the industrial computer (9) for signal transmission; The automatic wood chip grading method of the system comprises the following steps: S1: Identify surface defects of wood chips and grade them; S2: Accurately locate the position of the wood chips: S3: conveying and controlling wood chips, and finally sorting the graded wood chips; Wherein said S1 comprises: S11: Acquire a black and white image of the wood chip; S12: Obtaining defect outer contour points of the black and white image according to the black and white image; S13: Acquire the defect contour of the image; S14: Calculate the defect area of ​​the image; S15: Determine the grade of wood chips; Specifically, S11 includes: the information of the wood chip image actually collected includes: color data of red, green and blue, which is described as a five-dimensional array Color_Image[i,j,r,g,b], where i represents the horizontal sequence number of the current pixel; j represents the vertical sequence number of the current pixel; r represents the red value of the pixel; g represents the green value of the pixel; and b represents the blue value of the pixel; Color screening the image information of the wood chip, completing the setting of multiple pixel points, and obtaining a black and white image, which is composed of pixel points in both horizontal and vertical directions: setting the critical values ​​Cr, Cg, and Cb of the red, green, and blue colors of the image, and judging the red, green, and blue colors of all pixel points in the image: if the red, green, and blue color values ​​of the pixel point are all higher than the critical values ​​Cr, Cg, and Cb of the red, green, and blue colors, then the red, green, and blue pixel values ​​of the point are all set to 0; otherwise, the red, green, and blue pixel values ​​of the point are all set to 1; The S13 is as follows: take the first pixel point N[i,j] whose rgb value is equal to 0 in the image array Color_Image as the current point, where i is the horizontal serial number of the pixel point and j is the vertical serial number of the pixel point; and record N[i,j] points in the outline array Outline[l,k,i,j], and judge the value of the pixel point in the order of pixel points N[i,j-1], N[i-1,j-1], N[i-1,j], N[i-1,j+1], N[i,j+1], N[i+1,j+1], N[i+1,j], N[i-1,j+1], where l is the serial number of the current outline, whose initial value is 0 and its value increases by 1 for each additional outline; k is the lth outline. The current serial number of the point, its initial value is 0, each time a point is recorded, the k value automatically increases by 1; i is the horizontal serial number of the pixel point, j is the vertical serial number of the pixel point, if the RGB values ​​of Color_Image[x,y,r,g,b] are all equal to 0, where x is the horizontal serial number of the above 8 pixels; y is the vertical serial number of the above 8 pixels, then record N[x,y] points in the outline array Outline[l,k,i,j], and set the RGB values ​​of Color_Image[x,y,r,g,b] at this point to 1; repeat the above steps until all RGB values ​​in the Color_Image array are equal to 1, and obtain the outline array Outline[l,k,i,j]; The S14 is as follows: since the outline of the defect on the wood chip surface is a convex polygon, the number of all points of Outline[l,k,i,j] is k, and k-1 triangles can be formed by connecting lines from Outline[l,1,i,j] to points other than Outline[l,2,i,j] and the end point; for each triangle, the lengths of the three sides are calculated: a, b, c, and the average perimeter p = (a+b+c) / 2, according to the formula: triangle area s = 1 / 2 power of p(pa)(pb)(pc); the areas of all k-1 triangles are counted to obtain the area S of the defect, and the number of defects Num = Num+1 and the sum of the defect areas S are calculated. sum =S sum +S; After completing steps S11-S14 above, find the point where the pixel value v is not zero in the three-dimensional array Color_Image again, and execute steps S11-14 until all the pixels in the Color_Image array are 1; finally, the number Num of all defects on the wood chip and the sum of all defect areas S are obtained. sum .

2. The method for automatic grading of wood chips according to claim 1, wherein: The guide plate sorting mechanisms (7) are provided with three, namely, a Class A guide plate sorting mechanism, a Class B guide plate sorting mechanism, and a Class C guide plate sorting mechanism; correspondingly, a Class A collection frame, a Class B collection frame, and a Class C collection frame are provided below the guide plate sorting mechanisms.

3. The method for automatic grading of wood chips according to claim 1, wherein: The S12 is: if the RGB values ​​of Color_Image[i-1,j,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i+1,j,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i,j-1,r,g,b] are all equal to 1, and the RGB values ​​of Color_Image[i,j+1,r,g,b] are all equal to 1, then the point is judged to be an internal point of the defect area, and the RGB values ​​of the point are set to 1, otherwise they are set to 0.

4. The method for automatic grading of wood chips according to claim 1, wherein: Said S15 is: preset the critical defect number N of grade A plate a , B-grade board critical defect number N b and critical defect area S of grade A plate a , B-grade plate critical defect area (S b ); When the number of defects on the wood chip Num <N a And the defect area S sum a It is judged as Grade A board; when the number of defects on the wood chip is N a <Num<N b And the defect area S a sum b Determined to be a Grade B board;​​​ When the number of defects on the wood chip Num>N b And the defect area S sum >S b It is judged to be a C-grade board.

5. The method for automatic grading of wood chips according to claim 1, characterized in that: The method of S2 is as follows: the distance between the pneumatic suction cup array (3) at the end of the robot arm and the wood piece is measured using a distance sensor (4), the distance sensor measures the propagation time t of the emitted sound wave or light wave between the pneumatic suction cup and the wood piece, and multiplies the propagation time t by the speed of the sound wave or light wave to obtain the distance between the wood piece and the pneumatic suction cup; The method of S3 is as follows: a first servo motor drives a first conveyor belt to move wood chips, and the wood chips pass through a detection chamber (6) and a guide plate sorting mechanism (7) in sequence during the conveyor belt conveyance process. After the industrial computer (9) determines the grade of the wood chips, it sends a motion instruction to the corresponding guide plate sorting mechanism (7) to rotate and guide the wood chips to fall into the corresponding wood chip collection frame (8), thereby completing the sorting operation, wherein the initial position of the guide plate sorting mechanism is in a horizontal state.

Images