A two-dimensional layout method and system for irregularly shaped defective leather

Abstract

This invention develops a two-dimensional layout method and system for irregularly shaped defective leather. The method is as follows: Step 1: Acquire an image of the irregularly shaped defective leather with the defect locations marked; Step 2: Construct a layout problem model for the irregularly shaped defective leather, and formulate a global optimization objective function and a local optimization objective function for the layout process; Step 3: Preprocess the image of the irregularly shaped defective leather using computer vision technology; Step 4: Based on the leather outline information and cutting information, sequentially extract the starting layout position for each search using the generated coordinate point layout sequence, and move the optimal position of the layout objective value and the type of cutting block according to the local objective function; Step 5: Use the coordinate point sequence as the taboo object in the taboo list, and perform taboo search to find the optimal coordinate point layout sequence. This invention reduces production costs, saves time, and improves product quality, bringing beneficial innovation to the leather manufacturing industry.

Description

Technical Field

[0001] This invention belongs to the field of irregular leather layout technology, specifically relating to a two-dimensional layout method and system for irregular and defective leather. Background Technology

[0002] Leather layout, a crucial link in the leather industry, has always been a focus of attention. Leather, as a precious raw material, is widely used in fashion, footwear, automotive interiors, and furniture, profoundly impacting product appearance, quality, and sustainability. For example, a domestic furniture company uses imported automated nesting equipment, incurring tens of thousands of yuan in annual maintenance costs. While domestic companies are continuously developing automated nesting equipment, this requires long-term technological investment. Currently, this is a weakness in the domestic industry. Traditional layout methods are primarily based on regular geometric shapes and mostly study scenarios where leather is arranged in matrices, requiring each piece of leather to be intact. However, this scenario is clearly unrealistic and fails to adequately adapt to the irregularity and diversity of leather raw materials. The problem of irregular leather layout (also known as nesting) falls under the category of cutting and packaging problems, involving allocating a set of irregular or regular components into a larger rectangular or irregular container while minimizing material or space waste. Summary of the Invention

[0003] In response to the problems existing in current research and practical applications, this invention designs a two-dimensional nesting method and system for irregularly defective leather, which can improve material utilization and work efficiency, reduce enterprise production costs, and solve the problems existing in the background technology.

[0004] The technical solution of the present invention is as follows:

[0005] A two-dimensional layout method for irregularly shaped, defective leather, specifically including the following steps:

[0006] Step 1: Obtain an irregularly shaped piece of leather with marked defect locations and image it using a line scan camera to obtain a leather image;

[0007] Step 2: Construct a layout problem model for irregularly shaped defective leather, and formulate the global optimization objective function and the local optimization objective function in the layout process;

[0008] Step 3: Preprocess the images of irregularly shaped, defective leather using computer vision technology;

[0009] Step 4: Based on the leather outline information and cutting piece information, extract the starting position of each search sequentially through the generated coordinate point layout sequence, and move the optimal position of the layout target value and the type of cutting block according to the local objective function;

[0010] Step 5: Treat the coordinate point sequence as a taboo object in the taboo list and perform taboo search to find the optimal coordinate point sorting sequence.

[0011] Furthermore, step 2 specifically involves:

[0012] The quantity of leather to be processed is M, M j This represents the set of non-zero grid coordinates for the j-th leather to be arranged. This represents the leather mass at the (x, y) coordinate of the j-th leather to be arranged. A value of 0 indicates a hole location. max That is the value with the largest mass. This represents the nth arrangement method for the j-th leather to be arranged. This represents the sequence of coordinate points corresponding to the nth arrangement method for the j-th leather to be arranged. This represents the sum of the masses of the positions already arranged in the nth arrangement of the j-th leather to be arranged. Let represent the coordinate set of the cut block in the nth layout method for the j-th leather to be laid out. This represents the number of layouts at each (x, y) position in the nth layout method for the j-th leather to be laid out;

[0013] The global optimization objective is:

[0014]

[0015] The constraints are:

[0016]

[0017] θ i ∈θ(0≤θ≤π) (4)

[0018] In the layout scheme, the rotation angle of the cut blocks is θ. i The allowable rotation angle of the leather to be laid out relative to its initial posture is θ (the unit rotation angle can be customized, for example, setting the unit rotation to θ). The angle of rotation can only be 0. Three methods are used to better utilize the available space of the leather to be laid out compared to not rotating or rotating only 90° or multiples thereof. The coordinate set of each piece of leather to be laid out in the final layout scheme is... The usable area corresponding to this plan is Z. n Equation (1) is the objective function of two-dimensional layout optimization, which mainly means maximizing the utilization rate of leather. Equation (2) indicates that the number of each layout can only be 0 or 1. Equation (3) indicates that the number of layouts corresponding to the coordinate point set in the layout scheme cannot be greater than 1, otherwise it indicates that there is an overlapping part. Equation (4) indicates the angle of rotation allowed for the leather to be arranged.

[0019] The local optimization objective function is:

[0020]

[0021] in, The weighted sum of each local optimization objective is used, where σ(x,y,θ) represents the total area of ​​the pre-laid blocks after a rotation angle θ at (x,y), μ(x,y,θ) represents the proportion of the pre-laid block area to the minimum bounding matrix of the pre-laid blocks after a rotation angle θ at (x,y), and β(x,y,θ) represents the sum of the number of un-laid positions above and to the left of the pre-laid block after a rotation angle θ at (x,y). Finally, k1, k2, and k3 (which balance the importance of the three optimization objectives) are used to weight the sum of each optimization objective.

[0022] Furthermore, step 3 specifically involves:

[0023] Step 3.1: Perform binary grayscale processing on the rasterized leather data.

[0024] Step 3.2: Using the Canny edge detection algorithm, save the pixels with values ​​greater than the set threshold (save the defect line traces), find other connected pixel regions besides the line, obtain the contour information of the line, and draw the contour information onto the image.

[0025] Step 3.3: Use the pre-set dual thresholds to outline all the defect locations on the irregular leather with the same color as the background.

[0026] Step 3.4: Use the Canny edge detection algorithm to find the outer contour information of the entire cowhide and draw it on the image with a color that is the opposite of the background color (e.g., draw black if the background is white, or draw white if the background is black).

[0027] Step 3.5: Traverse the image information and process the interior of the drawn line (defect area) to be the same color as the background.

[0028] Furthermore, step 4 specifically involves:

[0029] Step 4.1: Extract the internal layout coordinate point sequence L based on the leather outline information, and compress the distance between the coordinate points according to the value of the minimum layout block.

[0030] Step 4.2: Extract the starting search point position P in sequence L, and initialize the optimal sorting position S corresponding to the current position P. max .

[0031] Step 4.3: Retrieve block i from the layout block list, obtain the type and size of the corresponding block, and initialize an optimal layout position for the current block. Let S be the initial search point position. p .

[0032] Step 4.4: Calculate the objective function value V corresponding to the current position in each unit direction based on the objective function obtained by weighting the three loss functions, and extract the maximum objective function value V in each direction. max .

[0033] Step 4.5: If the maximum objective function value V is obtained... max The objective function value V at the current position is less than or equal to the objective function value V. local If the local search for the current cut block terminates, proceed to step 4.6; if the optimal nesting position corresponds to the current position P. Then update S max for Return to step 4.3.

[0034] Step 4.6: Update the search point position P to the position after moving one unit in the direction of the maximum target value. At the same time, use perspective transformation technology to determine whether there are other cut blocks already laid out at the current position or whether other laying out requirements are not met. If the current position is feasible, update... And return to step 4.4; otherwise, keep constant.

[0035] Furthermore, step 5 specifically involves:

[0036] Step 5.1: Define each layout scheme, i.e. the initial search coordinate sequence L of each layout, as a taboo object. The taboo length is the taboo table length. Generate an initial solution from left to right and top to bottom based on the internal coordinate point set of the leather.

[0037] Step 5.2: Define the neighborhood movement rule as randomly shuffling the search coordinate sequence, generating multiple candidate solutions from the current optimal solution, calculating the global objective function of each candidate solution, and selecting the objective value from the candidate solutions. The largest candidate solution is added to the tabu list; if the tabu list is full, the head of the tabu list is removed.

[0038] Step 5.3: Compare the best candidate solution with the current best solution. If the best candidate solution is better than the current best solution, update the current best candidate solution; otherwise, keep the current best solution.

[0039] Step 5.4: If the maximum number of iterations is reached or the optimal solution cannot be updated within a given number of iterations, then the search is terminated.

[0040] This invention also discloses a two-dimensional nesting system for irregularly shaped defective leather, based on the above method, which includes the following modules:

[0041] Image acquisition module: Acquires images of irregularly shaped leather with defects, with the defect locations marked;

[0042] Objective function formulation module: Construct a model for the layout problem of irregular and defective leather, and formulate the global optimization objective function and the local optimization objective function for the layout process;

[0043] Image preprocessing module: Preprocesses images of irregularly shaped, defective leather using computer vision technology;

[0044] The layout position module: Based on the leather outline information and cutting piece information, it sequentially extracts the starting layout position for each search through the generated coordinate point layout sequence, and moves the layout target value to the optimal position and the type of cutting piece according to the local objective function;

[0045] Search module: Treats the coordinate point sequence as taboo objects in the taboo list and performs taboo search to find the optimal coordinate point sorting sequence.

[0046] The beneficial effects of this invention are as follows: By utilizing computer vision technology and optimization algorithms, leather can be effectively laid out, minimizing waste and improving leather utilization, thereby reducing production costs. Traditional leather layout typically requires a large amount of manual operation, while this invention automates the layout process using computer vision and algorithms, reducing reliance on manual labor and saving labor costs. Through rasterized representation and mathematical modeling of the imaging system, this invention can more accurately identify and locate leather defects, making layout more precise and improving layout accuracy, avoiding mishandling of defects during production. This invention employs optimization algorithms such as tabu search, which can generate and continuously optimize layout schemes in a short time, improving layout efficiency, saving time, and shortening the production cycle. By preprocessing digitally defective leather, this invention can better identify and handle defects, thereby improving the quality and appearance of the final product. Attached Figure Description

[0047] Figure 1 This is a flowchart of a two-dimensional layout method for irregularly shaped defective leather, according to a preferred embodiment.

[0048] Figure 2 This is a preprocessing result diagram of a two-dimensional nesting method for irregularly shaped defective leather according to a preferred embodiment.

[0049] Figure 3 This is a preferred embodiment of a two-dimensional nesting method for irregularly shaped defective leather, which initializes the sequence coordinate diagram.

[0050] Figure 4 This is a preferred embodiment of a two-dimensional layout method for irregular, defective leather, and the layout effect diagram is shown.

[0051] Figure 5 This is a block diagram of a two-dimensional layout system for irregularly shaped, defective leather, according to a preferred embodiment. Detailed Implementation

[0052] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0053] Reference Figure 1-4 This embodiment provides a two-dimensional layout method for irregularly shaped defective leather, comprising the following steps:

[0054] Step 1: Obtain an irregularly shaped piece of leather with defects marked by workers and image it using a line scan camera;

[0055] Step 2: Construct a model for the irregular and defective cowhide layout problem, and formulate the global optimization objective function and the local optimization objective function in the layout process;

[0056] Step 3: Preprocess the images of irregularly shaped, defective leather using computer vision technology;

[0057] Step 4: Based on the leather outline information and cutting piece information, extract the starting position of each search sequentially through the generated coordinate point layout sequence, and move and layout to the optimal position and type of cutting block according to the local objective function;

[0058] Step 5: Treat the coordinate point sequence as the taboo object in the taboo list and perform taboo search to find the optimal coordinate point sorting sequence;

[0059] In this embodiment, step 2 specifically includes:

[0060] The quantity of leather to be processed is M, M j This represents the set of non-zero grid coordinates for the j-th leather to be arranged. This represents the leather mass at the (x, y) coordinate of the j-th leather to be arranged. A value of 0 indicates a hole location. max That is the value with the largest mass. This represents the nth arrangement method for the j-th leather to be arranged. This represents the sequence of coordinate points corresponding to the nth arrangement method for the j-th leather to be arranged. This represents the sum of the masses of the positions already arranged in the nth arrangement of the j-th leather to be arranged. Let represent the coordinate set of the cut block in the nth layout method for the j-th leather to be laid out. This represents the number of layouts at each (x, y) position in the nth layout method for the j-th leather to be laid out;

[0061] The global optimization objective is:

[0062]

[0063] The constraints are:

[0064]

[0065] θ i ∈θ(0≤θ≤π) (4)

[0066] In the layout scheme, the rotation angle of the cut blocks is θ. i The allowable rotation angle of the leather to be laid out relative to its initial posture is θ (the unit rotation angle can be customized, for example, set to θ). The angle of rotation can only be 0. Three methods are used to better utilize the available space of the leather to be laid out compared to not rotating or rotating only 90° or multiples thereof. The coordinate set of each piece of leather to be laid out in the final layout scheme is... The usable area corresponding to this plan is Z. n Equation (1) is the objective function of two-dimensional layout optimization, which mainly means maximizing the utilization rate of leather. Equation (2) indicates that the number of each layout can only be 0 or 1. Equation (3) indicates that the number of layouts corresponding to the coordinate point set in the layout scheme cannot be greater than 1, otherwise it indicates that there is an overlapping part. Equation (4) indicates the angle of rotation allowed for the leather to be arranged.

[0067] The local optimization objective function is:

[0068]

[0069] in, The weighted sum is calculated for each local optimization objective. σ(x,y,θ) represents the total area of ​​the pre-laid blocks after a rotation angle θ at (x,y), μ(x,y,θ) represents the proportion of the pre-laid block area to the minimum bounding matrix of the pre-laid blocks after a rotation angle θ at (x,y), and β(x,y,θ) represents the sum of the number of un-laid positions above and to the left of the pre-laid block after a rotation angle θ at (x,y). Finally, k1, k2, and k3 are used to weight the sum for each optimization objective.

[0070] In this embodiment, step 3 specifically includes:

[0071] Step 3.1: Perform binary grayscale processing on the rasterized leather data.

[0072] Step 3.2: Using the Canny edge detection algorithm, save the pixels with values ​​greater than the set threshold (save the defect line traces), find other connected pixel regions besides the line, obtain the contour information of the line, and draw the contour information onto the image.

[0073] Step 3.3: Use the pre-set dual thresholds to outline all the defect locations on the irregular leather with the same color as the background.

[0074] Step 3.4: Use the Canny edge detection algorithm to find the outer contour information of the entire cowhide, and draw it on the image with a color that is the opposite of the background color (e.g., if the background is white, draw black, or if the background is black, draw white).

[0075] Step 3.5: Traverse the image information and process the interior of the drawn line (defect area) to the same color as the background according to the algorithm.

[0076] In this embodiment, step 4 specifically includes:

[0077] Step 4.1: Extract the internal layout coordinate point sequence L based on the leather outline information, and compress the distance between the coordinate points according to the value of the minimum layout block.

[0078] Step 4.2: Extract the starting search point position P in sequence L, and initialize the optimal sorting position S corresponding to the current position P. max .

[0079] Step 4.3: Retrieve block i from the layout block list, obtain the type and size of the corresponding block, and initialize an optimal layout position for the current block. Let S be the initial search point position. p .

[0080] Step 4.4: Calculate the objective function value V corresponding to the current position in each unit direction based on the objective function obtained by weighting the three loss functions, and extract the maximum objective function value V in each direction. max .

[0081] Step 4.5: If the maximum objective function value V is obtained... max The objective function value V at the current position is less than or equal to the objective function value V. local If the local search for the current cut block terminates, proceed to step 4.6; if the optimal nesting position corresponds to the current position P. Then update S max for Return to step 4.3.

[0082] Step 4.6: Update the search point position P to the position after moving one unit in the direction of the maximum target value. At the same time, use perspective transformation technology to determine whether there are other cut blocks already laid out at the current position or whether other laying out requirements are not met. If the current position is feasible, update... And return to step 4.4; otherwise, keep constant.

[0083] In this embodiment, step 5 specifically includes:

[0084] Step 5.1: Define each layout scheme, i.e. the initial search coordinate sequence L of each layout, as a taboo object. The taboo length is the taboo table length. Generate an initial solution from left to right and top to bottom based on the internal coordinate point set of the leather.

[0085] Step 5.2: Define the neighborhood movement rule as randomly shuffling the search coordinate sequence, generating multiple candidate solutions from the current optimal solution, calculating the global objective function of each candidate solution, and selecting the objective value from the candidate solutions. The largest candidate solution is added to the tabu list; if the tabu list is full, the head of the tabu list is removed.

[0086] Step 5.3: Compare the best candidate solution with the current best solution. If the best candidate solution is better than the current best solution, update the current best candidate solution; otherwise, keep the current best solution.

[0087] Step 5.4: If the maximum number of iterations is reached or the optimal solution cannot be updated within a given number of iterations, then the search is terminated.

[0088] Reference Figure 3 This paper presents a novel two-dimensional layout method for irregularly shaped, defective leather and a system initialization sequence coordinate diagram. Through... Figure 3 It is understood that the present invention uses the outer contour information of the leather to be laid out and the minimum side length of the block to be laid out to uniformly divide the internal pixels of the leather, thereby obtaining the initial sequence of each block arrangement in the layout scheme.

[0089] Reference Figure 4 This paper demonstrates a novel two-dimensional layout method for irregularly shaped, defective leather, along with a diagram illustrating the layout effect. Through... Figure 4 It can be seen that the layout effect diagram of the present invention can obtain a high-utilization rotation layout scheme on defective leather to be laid out.

[0090] like Figure 5 As shown, this embodiment discloses a two-dimensional nesting system for irregularly shaped defective leather, which, based on the above method embodiment, includes the following modules:

[0091] Image acquisition module: Acquires images of irregularly shaped leather with defects, with the defect locations marked;

[0092] Objective function formulation module: Construct a model for the layout problem of irregular and defective leather, and formulate the global optimization objective function and the local optimization objective function for the layout process;

[0093] Image preprocessing module: Preprocesses images of irregularly shaped, defective leather using computer vision technology;

[0094] The layout position module: Based on the leather outline information and cutting piece information, it sequentially extracts the starting layout position for each search through the generated coordinate point layout sequence, and moves the layout target value to the optimal position and the type of cutting piece according to the local objective function;

[0095] Search module: Treats the coordinate point sequence as taboo objects in the taboo list and performs taboo search to find the optimal coordinate point sorting sequence.

[0096] Other aspects of this embodiment can be found in the above method embodiments.

[0097] In summary, this invention develops a novel two-dimensional nesting method and system for irregularly shaped defective leather. It performs two-dimensional nesting for irregularly shaped defective leather, achieving efficient nesting scheme generation through imaging system to mark the defect locations, mathematical modeling, and computer vision preprocessing. Its optimization objective is to maximize leather utilization while considering constraints such as rotation angles. This invention utilizes a tabu search algorithm to optimize the nesting scheme, improving nesting accuracy and efficiency. By digitally processing defects and using grayscale processing, leather defects are effectively identified and addressed, ultimately improving product quality. This invention can reduce production costs, save time, and improve product quality, bringing beneficial innovation to the leather manufacturing industry.

[0098] The above description is a preferred embodiment of the present invention. For those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A two-dimensional layout method for irregularly shaped, defective leather, characterized in that, Includes the following steps: Step 1: Obtain an image of the irregularly shaped leather with defects marked on it; Step 2: Construct a model for the layout problem of irregularly shaped defective leather, and formulate the global optimization objective function and the local optimization objective function for the layout process; Step 3: Preprocess the images of irregularly shaped, defective leather using computer vision technology; Step 4: Based on the leather outline information and cutting piece information, extract the starting position of each search sequentially through the generated coordinate point layout sequence, and move the optimal position of the layout target value and the type of cutting block according to the local objective function; Step 5: Treat the coordinate point sequence as the taboo object in the taboo list and perform taboo search to find the optimal coordinate point sorting sequence; Step 2 is as follows: The number of leathers to be arranged is M , Indicates the first j A set of non-zero grid coordinates for the leather to be arranged. Indicates the first j A leather product awaiting arrangement ( x,y The coordinates indicate the leather quality; a value of 0 indicates a hole location. That is the value with the largest mass. Indicates the first j The first batch of leather awaiting arrangement n Various sorting methods Indicates the first j The first batch of leather awaiting arrangement n The sequence of coordinate points corresponding to the various nesting methods This indicates the j-th leather to be arranged. n The mass of the positions already arranged in the various layout methods, This indicates the j-th leather to be arranged. n The coordinate set of the cut blocks in various layout methods Indicates the first j The first batch of leather awaiting arrangement n Each of the sorting methods The number of layouts at each location; The global optimization objective is: （1） in, N Indicates for the first j Number of leather layout methods; The constraints are: （2） （3） （4） In the layout scheme, the rotation angle of the cut blocks is: The allowable angle of rotation for the leather to be sampled relative to its initial position is... In the final layout scheme, the coordinate set of each leather block to be laid out is: The usable area corresponding to this plan is Equation (1) is the objective function of two-dimensional layout optimization, which means maximizing the utilization rate of leather. Equation (2) means that the number of each layout can only be 0 or 1. Equation (3) means that the number of layouts corresponding to the coordinate point set of the layout scheme cannot be greater than 1, otherwise it indicates that there is an overlapping part. Equation (4) means the angle of rotation allowed for the leather to be laid out. The local optimization objective function is: （5） in, The weighted sum of the local optimization objectives. Indicates in The rotation angle of the nesting is The total area of ​​the cut blocks that have been laid out is now complete. Indicates in The rotation angle of the nesting is The proportion of the area of ​​the laid-out cut pieces to the minimum bounding matrix of the laid-out cut pieces. Indicates in The layout rotation angle is the sum of the number of unplaced positions above and to the left of the cut piece; finally, use... The various optimization objectives are weighted and summed.

2. The two-dimensional layout method for irregularly shaped defective leather according to claim 1, characterized in that, Step 3 is as follows: Step 3.1: Perform binary grayscale processing on the rasterized leather data; Step 3.2: Using the Canny edge detection algorithm, save the pixels with values ​​greater than the set threshold, find the connected regions of other pixels except for the drawn line, obtain the contour information of the drawn line, and draw the contour information onto the image. Step 3.3: Using the set dual thresholds, trace all the defect locations of the irregular leather to the same color as the background; Step 3.4: Use the Canny edge detection algorithm to find the outer contour information of the entire leather piece, and draw it on the image with a different color than the background color; Step 3.5: Traverse the image information and process the internal defects at the drawn lines to be the same color as the background.

3. A two-dimensional nesting method for irregularly shaped defective leather according to any one of claims 1-2, characterized in that, Step 4 is as follows: Step 4.1: Extract the internal energy layout coordinate point sequence based on the leather contour information. L And compress the distance between coordinate points according to the value of the minimum cut block in the layout; Step 4.2: Extract the starting search point positions in sequence L. P and initialize the current position. P Corresponding optimal nesting position ; Step 4.3: Take out the cutting blocks from the pattern cutting block list. i It obtains the type and size of the corresponding cut block and initializes an optimal nesting position for the current cut block. Initial search point location ; Step 4.4: Calculate the objective function value corresponding to the current position in each unit direction based on the objective function obtained by weighting the three loss functions. V And extract the maximum objective function value in each direction. ; Step 4.5: If the maximum objective function value is obtained... The objective function value is less than or equal to the current position. If the current position is not found, the local search for the current cut block terminates, and the process jumps to step 4.6; if the current position is not found, the search for the current cut block terminates, and the process jumps to step 4.

6. P Corresponding optimal nesting position Then update for Return to step 4.3; Step 4.6: Update the search point position P to the position after moving it one set unit in the direction of the maximum target value. Use perspective transformation technology to determine if there are other cut blocks already laid out at the current position or if other layout requirements are not met. If so, update... If yes, return to step 4.4; otherwise, keep... constant.

4. The two-dimensional layout method for irregularly shaped defective leather according to claim 3, characterized in that, Step 5 is as follows: Step 5.1: Define each layout scheme, i.e. the initial search coordinate sequence L of each layout, as the taboo object. The taboo length is the taboo table length. Generate an initial solution from left to right and from top to bottom based on the internal coordinate point set of the leather. Step 5.2: Define the neighborhood movement rule as randomly shuffling the search coordinate sequence, generating multiple candidate solutions from the current optimal solution, calculating the global objective function of each candidate solution, and selecting the objective value from the candidate solutions. The largest candidate solution is added to the tabu list. If the tabu list is full, the head of the tabu list is removed. Step 5.3: Compare the best candidate solution with the current best solution. If the best candidate solution is better than the current best solution, update the current best candidate solution; otherwise, keep the current best solution. Step 5.4: If the maximum number of iterations is reached or the optimal solution cannot be updated within a given number of iterations, then the search is terminated.

5. A two-dimensional patterning system for irregularly shaped, defective leather, based on the method described in any one of claims 1-4, characterized in that, Includes the following modules: Image acquisition module: Acquires images of irregularly shaped leather with defects, with the defect locations marked; Objective function formulation module: Construct a model for the layout problem of irregular and defective leather, and formulate the global optimization objective function and the local optimization objective function for the layout process; Image preprocessing module: Preprocesses images of irregularly shaped, defective leather using computer vision technology; The layout position module: Based on the leather outline information and cutting piece information, it sequentially extracts the starting layout position for each search through the generated coordinate point layout sequence, and moves the layout target value to the optimal position and the type of cutting piece according to the local objective function; Search module: Treats the coordinate point sequence as taboo objects in the taboo list and performs taboo search to find the optimal coordinate point sorting sequence.

Images