Ton bag unloading method based on laser cutting

Abstract

The present application relates to the technical field of image data processing, and particularly relates to a ton bag unloading method based on laser cutting. First, a standard pixel matrix containing outer bag bottom image information and inner bag bottom image information is obtained. Then, an outer bag prediction frame and a ligature prediction frame are marked in the standard pixel matrix. Next, a safe offset distance is set. Then, the outer bag prediction frame is reduced by the safe offset distance to form an outer bag cutting frame; at the same time, the ligature prediction frame is expanded by the safe offset distance to form a ligature cutting frame; and the area enclosed between the outer bag cutting frame and the ligature cutting frame is defined as a cutting area. Finally, a cutting path is selected in the cutting area, and the laser cuts the outer bag bottom and the inner bag bottom along the cutting path for unloading. The present application accurately judges the position of the inner bag bottom ligature, and effectively improves the cutting accuracy and reduces the probability of ton bag debris polluting raw materials based on the high-temperature cutting of the laser.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image data processing, and particularly relates to a ton bag unloading method based on laser cutting. BACKGROUND

[0002] In some industries, ton bags with inner and outer double-layer structures are generally used for transporting granular materials. The traditional ton bag unloading step after transportation to the material yard is as follows: first, workers hang the ton bags on the bag hooks of the traveling crane one by one, the traveling crane is operated to the material pool or pit, and the workers cut the bottom of the ton bag, then the ton bag is lifted by the bag hook, and the ton bag is unloaded into the pit by relying on the gravity of the ton bag itself. Then, when the unloading is completed, the traveling crane is operated to the empty position, and the empty ton bag is removed by the workers at the empty position. Finally, the traveling crane is operated to the material taking position to start the next cycle.

[0003] Through the above unloading method, the ton bag unloading work can be effectively completed. However, in the unloading process, a large number of workers are involved, which leads to excessive occupation of human resources and high cost of manual unloading operation. At the same time, when the inner bag is used in the ton bag, the bottom of the inner bag has a ligation, which makes it easy for the worker to cut the ligation when using the cutter to cut, thereby causing the cutting to be blocked, and the cutting efficiency is reduced.

[0004] Therefore, in order to solve the above technical problems, most unloading sites introduce a mechanical ton bag crushing method to reduce the number of workers. Patent CN113562588B discloses a mechanical cutting device, which cuts the ton bag hung by the bag hook from the side without ligation by the cutter in the cutting assembly, thereby realizing the function of automatic unloading without the assistance of workers and saving labor costs. However, when the mechanical cutter cuts the ton bag, the physical tearing of the cutter to the polymer material inevitably produces debris, which mixes into the raw material and causes pollution of the raw material, thereby affecting the smooth progress of the subsequent processing flow of the raw material.

[0005] Therefore, the ton bag cutting method used at the present stage still needs to be further improved to improve the ton bag cutting effect. SUMMARY

[0006] In order to avoid and overcome the technical problems existing in the prior art, the present application provides a ton bag unloading method based on laser cutting. The present application accurately judges the position of the inner bag bottom ligation, and effectively improves the cutting accuracy and reduces the probability of ton bag debris polluting the raw material based on the high-temperature cutting of the laser.

[0007] To achieve the above purpose, the present application provides the following technical scheme:

[0008] A ton bag unloading method based on laser cutting, comprising the following unloading steps:

[0009] S1, obtaining a standard pixel matrix containing image information of the bottom of the outer bag and image information of the bottom of the inner bag when the bag hook hoists the ton bag to the unloading place;

[0010] S2, marking an outer bag prediction frame framing the bottom of the outer bag and a ligation prediction frame framing the ligation at the bottom of the inner bag in the standard pixel matrix;

[0011] S3, setting a safe offset distance between the outer bag prediction frame and the ligation prediction frame according to the relative position between them;

[0012] S4, forming an outer bag cutting frame by reducing the outer bag prediction frame by the safe offset distance; forming a ligation cutting frame by expanding the ligation prediction frame by the safe offset distance; and defining the area enclosed between the outer bag cutting frame and the ligation cutting frame as a cutting area;

[0013] S5, selecting a cutting path in the cutting area, and making the laser cut the bottom of the outer bag and the bottom of the inner bag along the cutting path.

[0014] As a further scheme of the present application, the process of obtaining the standard pixel matrix is as follows:

[0015] S11, when the ton bag is hoisted to the unloading place by the bag hook, the hyperspectral camera located below the ton bag is started;

[0016] S12, turn on the coaxial SWIR light source, generate a three-dimensional data cube by single shooting of the hyperspectral camera, and extract the original pixel matrix of the spatial dimension :

[0017] S13, convert the original pixel matrix into a standard pixel matrix by normalization operation.

[0018] As a further scheme of the present application, the original pixel matrix is represented as follows:

[0019] (1);

[0020] Wherein, represents the original pixel value at the intersection of the th row and the th column of the original pixel matrix, , , wherein and represent the total number of rows and the total number of columns of the original pixel matrix, respectively.

[0021] As a further scheme of the present application, the normalization operation is represented as follows:

[0022] (2);

[0023] wherein, represents the normalized standard pixel value; represents the pixel mean value in the original pixel matrix; represents the pixel standard deviation in the original pixel matrix.

[0024] As a further scheme of the present application: the labeling process of the prediction frame is as follows:

[0025] S21, labeling the outer bag prediction frame of the outer bag bottom part and the ligature prediction frame of the inner bag bottom part in the standard pixel matrix ; wherein represents the coordinate of the center point of the outer bag bottom part in the pixel coordinate system, and respectively represent the height value and the width value of the outer bag prediction frame, represents the coordinate of the center point of the inner bag bottom part ligature in the pixel coordinate system, and respectively represent the height value and the width value of the ligature prediction frame;

[0026] S22, inputting the standard pixel matrix labeled with the outer bag prediction frame and the ligature prediction frame into the target detection algorithm to train a target prediction model;

[0027] S23, using the target prediction model to predict the outer bag prediction frame and the ligature prediction frame in the new standard pixel matrix.

[0028] As a further scheme of the present application: when the cutting path is too far away from the ton bag edge, the uncut area will form an L-shaped stagnant space due to the self-weight of the material, causing part of the material to remain in the space, affecting the thoroughness of unloading; when the cutting path is too close to the ton bag edge, the cutting edge is prone to fracture or discontinuity due to laser thermal stress concentration, and the weak cutting edge is difficult to withstand the impact of the material, which may cause leakage risk. Therefore, a moderate safety offset distance needs to be determined by fusing the dynamic algorithm of the material characteristics (such as density, repose angle) and the ton bag structure parameters (such as size) to balance the unloading effect and cutting stability. The safety distance is represented as follows:

[0029] (3);

[0030] wherein, represents the allowable residual material mass in the ton bag; represents the repose angle of the material in the ton bag, represents the density of the material in the ton bag, ​Scaling factor for pixel coordinate system of hyperspectral camera Indicates the tangent function.

[0031] As a further scheme of the present application: the acquisition process of the cutting area is as follows:

[0032] S41, taking the center point of the outer bag prediction frame as the scaling base point, the height and width of the outer bag prediction frame are reduced by a safe offset distance, to form an outer bag cutting frame , and the standard pixel matrix at this time is valued: when the pixel point is located on the outer bag cutting frame, the pixel value of the pixel point is taken as 2; when the pixel point is located in the outer bag cutting frame, the pixel value of the pixel point is taken as 1; when the pixel point is located outside the outer bag cutting frame, the pixel value of the pixel point is taken as 0; the standard pixel matrix after the value assignment at this time is recorded as the outer bag pixel matrix ;

[0033] In another standard pixel matrix, taking the center point of the ligation prediction frame as the scaling base point, the height and width of the ligation prediction frame are expanded by a safe offset distance, to obtain a ligation cutting frame , and the standard pixel matrix at this time is valued: when the pixel point is located on the ligation cutting frame, the pixel value of the pixel point is taken as 3; when the pixel point is located in the ligation cutting frame, the pixel value of the pixel point is taken as 0; when the pixel point is located outside the ligation cutting frame, the pixel value of the pixel point is taken as 1; the standard pixel matrix after the value assignment at this time is recorded as the ligation pixel matrix ;

[0034] S42, the outer bag pixel matrix and the ligation pixel matrix are multiplied to form a superimposed pixel matrix ;

[0035] S43, the relative position of the outer bag cutting frame and the ligation cutting frame has two typical scenarios: the ligation cutting frame is completely located inside the outer bag cutting frame, then the square ring area between the two needs to be cut; the two cutting frames partially intersect, then the U-shaped area outside the overlapping area of the two needs to be cut.

[0036] Because the definition rules of the cutting areas corresponding to the two scenarios are different, a quantitative index is needed to distinguish, and this index is defined as the interference coefficient :

[0037] (4);

[0038] (5);

[0039] (6);

[0040] (7);

[0041] (8);

[0042] wherein, , , and represent transition parameters;

[0043] S44, when the ligature cutting frame is located inside the outer bag cutting frame, i.e. , the cutting path matrix :

[0044] (9);

[0045] When the ligature cutting frame intersects with the outer bag cutting frame, i.e. , the cutting path matrix :

[0046] (10);

[0047] wherein, represents the pixel value of the intersection of the row and the column in the cutting path matrix; represents the pixel value of the intersection of the row and the column in the superimposed pixel matrix.

[0048] The area formed by the combination of the pixel points with pixel values of 1 in the cutting path matrix constitutes the cutting area.

[0049] As can be seen from formulas (3) to (10), the calculation of the interference coefficient depends on the boundary parameters of the outer bag cutting frame and the ligature cutting frame, and the spatial overlap range of the two frames is quantified by defining four transition parameters:

[0050] , the maximum value of the left boundaries of the two frames represents the left boundary of the overlap area.

[0051] , the minimum value of the right boundaries of the two frames represents the right boundary of the overlap area.

[0052] , the maximum value of the upper boundaries of the two frames represents the upper boundary of the overlap area.

[0053] , the minimum value of the lower boundaries of the two frames represents the lower boundary of the overlap area.

[0054] Through the four parameters, the width R-L and height B-T of the overlapping area of the two frames can be calculated, and then the area (R-L) x (B-T) of the overlapping area, the total area (H1+D) x (W1+D) of the ligature cutting frame after expanding the safety offset distance D.

[0055] When (R-L) x (B-T) > 0, it means that R > L (there is an overlapping width in the horizontal direction) and B > T (there is an overlapping height in the vertical direction), that is, the outer bag cutting frame and the ligature cutting frame actually overlap in space, that is, the ligature cutting frame is inside the outer bag cutting frame or the ligature cutting frame intersects with the outer bag cutting frame.

[0056] When (R-L) x (B-T) ≤ 0, it means that the outer bag cutting frame and the ligature cutting frame do not overlap in space, and they are in a completely separated state, that is, the ligature cutting frame is outside the outer bag cutting frame.

[0057] The area ratio refers to the ratio of the total area of the ligature cutting frame after expansion to the overlapping area of the outer bag cutting frame and the ligature cutting frame, which quantifies the spatial position relationship of the two cutting frames: when the area ratio S > 1, it means that the total area of the ligature cutting frame is greater than the overlapping area, that is, the ligature cutting frame is completely inside the outer bag cutting frame; when the area ratio S ≤ 1, it means that the total area of the ligature cutting frame is less than or equal to the overlapping area, that is, the ligature cutting frame partially intersects with the outer bag cutting frame or the edges are tangent.

[0058] Through this ratio, the position scenarios of the two frames can be accurately distinguished, providing a basis for subsequent dynamic definition of cutting areas (such as square ring areas or U-shaped areas), ensuring that the cutting path avoids ligature obstacles and covers the effective unloading area.

[0059] The relative position of the outer bag and the inner bag ligature varies due to different specifications and loading conditions. The interference coefficient realizes dynamic adaptation to complex structures through quantitative analysis, avoiding excessive cutting area (residual material) or insufficient cutting area (incomplete cutting) caused by incorrect position relationship judgment, which is a key link in path planning, ensuring that the laser cutting can efficiently open the ton bag and avoid ligature and other obstacles.

[0060] As a further scheme of the present application: the outer bag cutting frame and the ligature cutting frame are both rectangular, and the corresponding edges are parallel to each other; when the ligature cutting frame is located within the outer bag cutting frame, a square ring-shaped cutting area is formed between the outer bag cutting frame and the ligature cutting frame, at this time a U-shaped cutting path spanning three edges is formed in the cutting area, the U-shaped cutting path is composed of continuous pixel points in the cutting path matrix, and the laser is cut along the cutting path to unload the outer bag bottom and the inner bag bottom; when the ligature cutting frame intersects with the outer bag cutting frame, a U-shaped cutting area is formed between the outer bag cutting frame and the ligature cutting frame, at this time a matching U-shaped cutting path is formed in the cutting area, the U-shaped cutting path is composed of continuous pixel points in the cutting path matrix, and the laser is cut along the cutting path to unload the outer bag bottom and the inner bag bottom.

[0061] As a further scheme of the present application: when the pixel value of the pixel point on the cutting path is less than the standard pixel value and greater than the standard pixel value, it is considered that the cutting path passes through the outer bag reinforcing rib, and the instantaneous output power of the laser cutting head during cutting of the outer bag is adjusted according to the laser power adjustment principle; wherein, is the median of the pixel value in the standard pixel matrix. As a further scheme of the present application: the laser power adjustment principle is specifically represented as follows:

[0062] (11);

[0063] In the formula, represents the laser cutting instantaneous output power at the intersection of the row and the column of the superimposed pixel matrix; is the initial output power of the laser cutting.

[0064] Compared with the prior art, the present application has the following advantages:

[0065] The present application adopts hyperspectral camera penetration imaging technology to accurately identify the inner bag ligature point, and combines a dynamic safety distance algorithm based on material characteristics to realize intelligent optimization of the cutting path; through laser fusion cutting process, the operation is carried out in a closed environment, effectively preventing the debris pollution caused by mechanical cutting. Therefore, the present application effectively solves the problem of material pollution in the traditional ton bag unpacking operation by integrating three core technologies of hyperspectral visual recognition, dynamic laser power adjustment and intelligent path planning. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 is the unloading flowchart of the present application.

[0067] Figure 2 is the schematic diagram of the outer bag cutting frame in the present application.

[0068] Figure 3 ​The figure is a schematic diagram of the ligation cutting frame in the present application.

[0069] Figure 4 The figure is a schematic diagram of the ligation cutting frame in the present application.

[0070] Figure 5 The figure is a schematic diagram of the ligation cutting frame in the present application. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0072] Please refer to Figure 1 In the embodiments of the present application, the ton bag adopts a double-layer polypropylene structure, the outer bag is 0.3 mm thick, and the inner bag is 0.2 mm thick. The material loaded in the ton bag is cement powder, the density of which is ρ=1.2 g / cm³, and the repose angle is θ=35°.

[0073] I. System configuration

[0074] 1. Hyperspectral imaging system

[0075] A suitable hyperspectral camera is selected to accurately obtain the three-dimensional data cube of the ton bag, which lays a good foundation for subsequent image analysis. At the same time, a corresponding light source needs to be set to irradiate the bottom of the ton bag to improve the brightness of the shooting scene and enhance the shooting effect.

[0076] 2. Laser cutting system:

[0077] A laser cutting head is connected to the fiber laser, the laser cutting head is driven by a two-axis linkage mechanical arm, and a laser power adjuster is configured.

[0078] The two-axis linkage mechanical arm can drive the laser cutting head to move in the horizontal plane to cut the ton bag according to the predetermined cutting path.

[0079] 3. Installation position:

[0080] A tilted transparent glass plate is installed at the unloading position, and the hyperspectral camera, light source, fiber laser, and cutting head are all installed below the glass plate, so that after laser cutting of the ton bag, the material falls on the glass plate and slides along the glass plate into the bin. At the same time, in order to ensure the cleanliness of the glass plate, a cleaning unit such as a wiper can be provided on the glass plate to clean the material on the glass plate in time, avoiding the influence of subsequent cutting work due to material shielding. Or no cleaning unit is set, and regular manual cleaning is used for cleaning.

[0081] II. Unloading

[0082] 1. Obtain standard pixel matrix

[0083] The bag hook hoists the ton bag to the unloading position, and the center point of the ton bag is on the same vertical line with the center point of the vertical projection of the glass plate, so as to ensure that the ton bag is located within the shooting range of the hyperspectral camera. The position sensor at the unloading position senses the ton bag and transmits a signal to the hyperspectral camera, and the hyperspectral camera starts exposure.

[0084] Turn on the coaxial SWIR light source, generate a three-dimensional data cube by single shooting of the hyperspectral camera, and extract the original pixel matrix with a spatial dimension of 512x512 :

[0085] ;

[0086] Calculate the normalization parameter, and obtain , ;

[0087] Perform normalization operation: calculate the corresponding pixel standard value by formula (2), and construct the corresponding standard pixel matrix :

[0088] .

[0089] 2. Standard prediction box

[0090] Label the outer bag prediction box in the standard pixel matrix, which frames the bottom of the outer bag And the ligation prediction box that frames the bottom of the inner bag In this embodiment, the outer bag prediction box is And the ligation prediction box is . Obtain 10,000 standard pixel matrices with labeled prediction boxes, input them into the YOLOv5 model as the target detection algorithm for training, to obtain the target prediction model. Then use the target prediction model to predict the outer bag prediction box and the ligation prediction box in the new standard pixel matrix.

[0091] 3. Calculate the safe offset distance

[0092] ;

[0093] First, taking the center point of the outer bag prediction frame as the scaling base point, the height and width of the outer bag prediction frame are both reduced by a safe offset distance inward to form an outer bag cutting frame . As shown in , the outermost square is the boundary frame of the entire standard pixel matrix, the middle square is the outer bag prediction frame, and the innermost square is the outer bag cutting frame. Figure 2

[0094] Then, taking the center point of the ligature prediction frame as the scaling base point, the height and width of the ligature prediction frame are both expanded by a safe offset distance outward to obtain a ligature cutting frame . As shown in , the outermost square is the boundary frame of the entire standard pixel matrix, the middle square is the ligature cutting frame, and the innermost square is the ligature prediction frame. Figure 3

[0095] 4. Generate the cutting area

[0096] The generated cutting frame is shown in Table 1.

[0097] Table 1 Outer bag cutting frame and ligature cutting frame

[0098] ;

[0099] Based on the obtained outer bag cutting frame and ligature cutting frame, the corresponding outer bag pixel matrix and ligature pixel matrix are obtained according to the corresponding matrix assignment rules.

[0100] The outer bag pixel matrix : when the pixel point is located on the outer bag cutting frame, the pixel value of the pixel point is taken as 2; when the pixel point is located within the outer bag cutting frame, the pixel value of the pixel point is taken as 1; when the pixel point is located outside the outer bag cutting frame, the pixel value of the pixel point is taken as 0; at this time, the standard pixel matrix after assignment is recorded as the outer bag pixel matrix.

[0101] The ligature pixel matrix : when the pixel point is located on the ligature cutting frame, the pixel value of the pixel point is taken as 3; when the pixel point is located within the ligature cutting frame, the pixel value of the pixel point is taken as 0; when the pixel point is located outside the ligature cutting frame, the pixel value of the pixel point is taken as 1; at this time, the standard pixel matrix after assignment is recorded as the ligature pixel matrix.

[0102] Then, the outer bag pixel matrix and the ligature pixel matrix are multiplied to form a superimposed pixel matrix ​​As for the pixel point , the pixel value in the outer bag pixel matrix is 1 within the outer bag cutting frame; the pixel value in the ligature pixel matrix is 3 on the ligature cutting frame; and the pixel value in the superimposed pixel matrix is 3.

[0103] The interference coefficient is calculated as follows:

[0104] ;

[0105] ;

[0106] ;

[0107] ;

[0108] Substituting equation (8) into equation (9), we have: ;

[0109] The cutting path matrix is generated based on equation (9). The region formed by the pixel points with a pixel value of 1 in the cutting path matrix constitutes the cutting region. Based on the data in Table 1 and the interference coefficient, it can be known that the outer bag prediction frame and the ligature prediction frame are both rectangular in the present example, and the corresponding edges are parallel to each other. As shown in FIG. 8, Figure 4 the outermost square frame is the boundary frame of the entire standard pixel matrix, the middle square frame is the outer bag cutting frame, and the innermost square frame is the ligature cutting frame. The ligature cutting frame is located within the outer bag cutting frame, and a square ring-shaped cutting region is formed between the outer bag cutting frame and the ligature cutting frame. At this time, a U-shaped cutting path (the U-shaped dashed line in FIG. 8) spanning three edges is formed in the cutting region. The U-shaped cutting path is composed of continuous pixel points in the cutting path matrix, and the laser is moved along the cutting path to cut the outer bag bottom and the inner bag bottom. Figure 4 Figure 4 As shown in FIG. 9, the outermost square frame is the boundary frame of the entire standard pixel matrix, the middle square frame is the outer bag cutting frame, and the innermost square frame is the ligature cutting frame. This is another form of cutting region that does not appear in the present embodiment. In

[0110] , the ligature cutting frame intersects with the outer bag cutting frame, and a U-shaped cutting region is formed between the outer bag cutting frame and the ligature cutting frame. At this time, a U-shaped cutting path (the U-shaped dashed line in FIG. 9) that is adapted to the U-shaped cutting region is formed in the cutting region. Figure 5 Figure 5 As shown in FIG. 10, Figure 5 the outermost square frame is the boundary frame of the entire standard pixel matrix, the middle square frame is the outer bag cutting frame, and the innermost square frame is the ligature cutting frame. This is another form of cutting region that does not appear in the present embodiment. In Figure 5 , the ligature cutting frame intersects with the outer bag cutting frame, and a U-shaped cutting region is formed between the outer bag cutting frame and the ligature cutting frame. At this time, a U-shaped cutting path (the U-shaped dashed line in FIG. 10) that is adapted to the U-shaped cutting region is formed in the cutting region.U-shaped dotted line in the U-shaped path, when laser cutting, from either end of the U-shaped dotted line to the other end along the dotted line can complete the cutting), the U-shaped cutting path is composed of consecutive pixel points in the cutting path matrix, and the laser is cut along the cutting path to the outer bag bottom and the inner bag bottom.

[0111] 5. Laser cutting execution

[0112] U-shaped path planning:

[0113] Two-axis linkage mechanical arm controls the laser cutting head to move along the U-shaped path in the U-shaped path in the U-shaped path, that is, to complete the cutting. Figure 4

[0114] Power dynamic adjustment:

[0115] When the outer bag reinforcing rib on the cutting path is detected, such as , the instantaneous output power of the laser cutting head when cutting the bag is adjusted , the instantaneous output power of the non-reinforcing rib area is 1000W, to complete the entire cutting process.

[0116] The above is only the preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art within the technical range disclosed by the present application, according to the technical scheme and the inventive concept of the present application, equivalent replacement or change, should be covered within the protection scope of the present application.​

Claims

1. A method of laser cutting based ton bag unloading, characterized by, The method comprises the following unloading steps: S1, when the ton bag is lifted to the unloading position by the bag hook, the standard pixel matrix containing the image information of the outer bag bottom and the image information of the inner bag bottom at this time is obtained; S2, the outer bag prediction box framing the outer bag bottom and the binding prediction box framing the binding position of the inner bag bottom are marked in the standard pixel matrix; S3, according to the relative position between the outer bag prediction box and the binding prediction box, the safe offset distance between the two is set; S4, the outer bag prediction box is reduced by the safe offset distance to form the outer bag cutting box; at the same time, the binding prediction box is expanded by the safe offset distance to form the binding cutting box; And the area enclosed between the outer bag cutting box and the binding cutting box is defined as the cutting area; S5, a cutting path is selected in the cutting area, and the laser cuts the outer bag bottom and the inner bag bottom along the cutting path for unloading; The standard pixel matrix is obtained as follows: S11, when the ton bag is lifted to the unloading position by the bag hook, the hyperspectral camera located below the ton bag is started; S12, turn on the coaxial SWIR light source, generate a three-dimensional data cube by single shooting of the hyperspectral camera, and extract the original pixel matrix of the spatial dimension : S13, transforming the original pixel matrix into a standard pixel matrix by a normalization operation S13, transforming the original pixel matrix into a standard pixel matrix by a normalization operation ; Original pixel matrix is represented as follows: ； wherein, represents the original pixel value at the intersection of the row and the column of the original pixel matrix, , wherein and represent the total number of rows and the total number of columns of the original pixel matrix, respectively; The normalization operation is shown as follows: ； wherein represents represents the standard pixel value after normalization; represents the mean value of the pixels in the original pixel matrix; represents the standard deviation of the pixels in the original pixel matrix.

2. A method of laser cutting based ton bag unloading as claimed in claim 1, wherein, The marking process of the prediction box is as follows: S21, marking the outer bag prediction box of the outer bag bottom in the standard pixel matrix and the ligation prediction box of the ligation of the inner bag bottom ; wherein represents the coordinates of the center point of the outer bag bottom in the pixel coordinate system, and respectively represent the height value and the width value of the outer bag prediction box, represents the coordinates of the center point of the ligation of the inner bag bottom in the pixel coordinate system, and respectively represent the height value and the width value of the ligation prediction box; S22, the standard pixel matrix marked with the outer bag prediction box and the binding prediction box is input into the target detection algorithm to train the target prediction model; S23, the target prediction model is used to predict the outer bag prediction box and the binding prediction box in the new standard pixel matrix.

3. A laser cutting based ton bag unloading method as claimed in claim 2, wherein, The safe distance is shown as follows: ； wherein denotes the mass of material allowed to remain in the ton bag; denotes the angle of repose of the material in the ton bag, denotes the density of the material in the ton bag, is a scaling factor for the pixel coordinate system of the hyperspectral camera; denotes the tangent function.

4. A method of laser cutting based ton bag unloading according to claim 3, wherein, The cutting area is obtained as follows: S41, taking the center point of the outer bag prediction frame as a scaling base point, reducing the height and width of the outer bag prediction frame by a safe offset distance to form an outer bag cutting frame , and for the standard pixel matrix at this time, when a pixel point is located on the outer bag cutting frame, the pixel value of the pixel point is taken as 2; when a pixel point is located in the outer bag cutting frame, the pixel value of the pixel point is taken as 1; when a pixel point is located outside the outer bag cutting frame, the pixel value of the pixel point is taken as 0; the standard pixel matrix after the assignment at this time is recorded as an outer bag pixel matrix ; In another standard pixel matrix, taking the center point of the ligation prediction frame as the scaling base point, the height and width of the ligation prediction frame are both enlarged outward by a safety offset distance to obtain a ligation cutting frame , and for the standard pixel matrix at this time, when a pixel point is located on the ligation cutting frame, the pixel value of the pixel point is taken as 3; when a pixel point is located in the ligation cutting frame, the pixel value of the pixel point is taken as 0; when a pixel point is located outside the ligation cutting frame, the pixel value of the pixel point is taken as 1; at this time, the standard pixel matrix after assignment is recorded as a ligation pixel matrix ​ S42, outer bag pixel matrix and ligating the pixel matrix performing a dot product to form a superimposed pixel matrix ; S43, calculate the interference coefficient based on the relative position between the outer bag cutting frame and the ligature cutting frame : ； ； ； ； ； wherein , , and all represent transition parameters; S44, when the ligating cutting frame is located inside the outer bag cutting frame, i.e. the cutting path matrix : ； When the ligating cutting frame intersects with the outer bag cutting frame, i.e. the cutting path matrix : ； wherein represents a pixel value of an intersection of the i-th row and the j-th column in the cutting path matrix; represents a pixel value of an intersection of the i-th row and the j-th column in the cutting path matrix; represents a pixel value of an intersection of the i-th row and the j-th column in the cutting path matrix; represents a pixel value of an intersection of the i-th row and the j-th column in the superimposed pixel matrix; represents a pixel value of an intersection of the i-th row and the j-th column in the superimposed pixel matrix; represents a The area formed by the pixel points with pixel value 1 in the cutting path matrix constitutes the cutting area.

5. A laser cutting based ton bag unloading method according to claim 4, characterized in that, The outer bag cutting box and the binding cutting box are both rectangles, and the corresponding edges are parallel to each other; when the binding cutting box is located in the outer bag cutting box, the cutting area between the outer bag cutting box and the binding cutting box is a square ring, at this time a U-shaped cutting path spanning three edges is formed in the cutting area, the U-shaped cutting path is composed of continuous pixel points in the cutting path matrix, and the laser cuts the outer bag bottom and the inner bag bottom along the cutting path for unloading; when the binding cutting box intersects with the outer bag cutting box, the cutting area between the outer bag cutting box and the binding cutting box is a U shape, at this time a matching U-shaped cutting path is formed in the cutting area, the U-shaped cutting path is composed of continuous pixel points in the cutting path matrix, and the laser cuts the outer bag bottom and the inner bag bottom along the cutting path for unloading.

6. A laser cutting based ton bag unloading method according to claim 5, characterized in that, When the pixel value of the pixel point on the cutting path is When the cutting path passes through the outer bag reinforcing rib, the instantaneous output power of the laser cutting head when cutting the bag is adjusted according to the laser power adjustment principle; wherein, is the median of the pixel values in the standard pixel matrix.

7. A laser cutting based ton bag unloading method according to claim 6, characterized in that, The laser power adjustment principle is shown as follows: ； In the formula, represents the instantaneous output power of the laser cutting at the intersection of the row and the column in the superimposed pixel matrix; is the initial output power of the laser cutting.

Images