A fully automatic edge banding equipment intelligent control method and system

Abstract

This invention provides an intelligent control method and system for a fully automatic edge banding equipment, belonging to the field of intelligent control technology. The method includes: performing median filtering on the super-resolution edge coordinates and sorting them clockwise to form an ordered set; calculating curvature markers to divide the edge segments; randomly selecting points as units to fit circles and optimizing them to obtain the final arc model; mapping the sampling points along the arc model back to the original coordinate system and merging them to form a fitted super-resolution edge coordinate set; acquiring the real-time status parameters of the edge banding equipment and normalizing the status parameters to obtain a correction factor for correcting measurement deviations; and correcting the fitted super-resolution edge coordinate set using the correction factor to obtain a corrected edge coordinate set. This invention achieves high-precision identification and edge banding control of the board edges, improving the accuracy and efficiency of edge banding.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to an intelligent control method and system for a fully automatic edge banding equipment for sheet metal. Background Technology

[0002] In the production of panel furniture, the quality of edge banding directly affects not only the aesthetics of the furniture but also its durability and environmental performance. Traditional panel edge banding equipment mainly uses mechanical or simple automated control methods. These methods have many shortcomings in practical applications and are difficult to meet the high precision and high efficiency requirements of modern furniture production.

[0003] For example, due to the complex and varied edge shapes of irregularly shaped boards, traditional equipment has difficulty in accurately controlling the degree of adhesion between the edge banding strip and the edge of the board, resulting in obvious gaps and black lines at the edge banding. This not only affects the aesthetics of the furniture, but also reduces the market competitiveness of the product. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent control method and system for a fully automatic edge banding equipment for sheet metal, which realizes high-precision identification and edge banding control of sheet metal edges, thereby improving the accuracy and efficiency of edge banding.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] In a first aspect, a fully automatic edge banding equipment for sheet metal is provided, the method comprising:

[0007] Step 1: Construct an octree structure and map pixel values ​​based on the original board image resolution and preset octree depth. Decompose the nodes through multiple layers of filters to obtain subband sets. Perform morphological processing on the high-frequency subbands to obtain a binarized edge map. After determining the edge points, connect them into curves using Hough transform and extract the super-resolution edge coordinates by interpolation.

[0008] Step 2: After median filtering of the super-resolution edge coordinates, sort them clockwise to form an ordered set. Calculate the curvature markers to divide the edge segments. Use each segment as a unit to randomly select points to fit a circle and optimize it to obtain the final arc model. Map the sampling points along the arc model back to the original image coordinate system and merge them to form the fitted super-resolution edge coordinate set.

[0009] Step 3: Obtain the real-time status parameters of the edge banding equipment, and normalize the status parameters to obtain the correction factor used to correct the measurement deviation.

[0010] Step 4: Correct the fitted super-resolution edge coordinate set using a correction factor to obtain the corrected edge coordinate set;

[0011] Step 5: Based on the corrected edge coordinate set, calculate the actual geometric dimensions of the board and the deviation of the actual geometric dimensions, and obtain the dimensional deviation score;

[0012] Step 6: Based on the dimensional deviation score, mark the boards that do not meet the requirements and control the equipment to perform the corresponding processing.

[0013] Furthermore, an octree structure is constructed based on the original board image resolution and a preset octree depth, and pixel values ​​are mapped. The nodes are then decomposed through multiple layers of filters to obtain a set of subbands, including:

[0014] Based on the resolution of the original board image and the preset octree depth, an octree structure is constructed, where each node in the octree structure represents a pixel block in the original board image.

[0015] The pixel values ​​of the original board image are mapped block by block to the nodes of the octree structure;

[0016] For each node of the octree, convolution operations are performed using a low-pass filter and a high-pass filter of Haar wavelet, respectively, generating one low-frequency subband and three high-frequency subbands after each decomposition.

[0017] The low-frequency subbands obtained from the current layer decomposition are used as input for the next layer decomposition. The data of the low-frequency subbands are redistributed to the nodes of the octree, and the node size is adjusted to match the resolution after the current decomposition. The filtering operation is repeated on the remapped low-frequency subbands to generate new low-frequency and high-frequency subbands. If the current layer number has not reached the preset layer number, the layer number is incremented and the filtering operation is repeated. If the preset layer number has been reached, the decomposition is terminated to obtain a set of subbands after multi-layer decomposition. The subband set includes the low-frequency subbands of the final layer and the high-frequency subbands of each layer.

[0018] Furthermore, a binarized edge map is obtained by morphological processing of the high-frequency subbands. After determining the edge points, Hough transform is used to connect them into a curve, and interpolation is used to extract the super-resolution edge coordinates, including:

[0019] Morphological processing is performed on the high-frequency subband, and edge points are determined by threshold segmentation in the morphologically processed high-frequency subband.

[0020] The edge points are input into the Hough transform to identify collinear points and fit them into a continuous curve. Adjacent curve segments are merged to form a continuous edge curve.

[0021] On a continuous edge curve, linear interpolation is used to obtain the interpolated coordinates of the edge points;

[0022] Extract the coordinates of the interpolated edge points to form a super-resolution edge coordinate set.

[0023] Furthermore, after median filtering of the super-resolution edge coordinates, they are sorted clockwise to form an ordered set. Curvature markers are then used to divide the edge segments, including:

[0024] Median filtering is applied to the super-resolution edge coordinates, and the filtered coordinate points are sorted in a clockwise direction according to the edge curve to form an ordered edge coordinate set.

[0025] The local curvature of ordered coordinate points is calculated using cubic spline interpolation. All points are traversed, and points whose curvature exceeds a threshold are marked as segmentation points.

[0026] The edge is divided into multiple continuous segments based on the dividing point.

[0027] Furthermore, points are randomly selected as units to fit circles and optimize them to obtain the final arc model. The sampled points along the arc model are mapped back to the original image coordinate system and merged to form the fitted super-resolution edge coordinate set, including:

[0028] Extract the set of coordinate points within each segment as an independent fitting unit;

[0029] Based on the set of coordinate points of each segment, three points are randomly selected from the coordinate points of the current segment, and the center and radius of the fitted circle are calculated. For all points within the segment, the distance from the current arc model is calculated.

[0030] Points with a distance less than the threshold are marked as interior points. The center and radius of the circle are re-optimized using the least squares method with all interior points to obtain the final arc model of each segment.

[0031] For the final circular arc model of each segment, sampling points are uniformly sampled along the circular arc path with a super-resolution step size;

[0032] The sampling points are mapped back to the original image coordinate system, and the dense coordinate points of all segments are combined in the original sorting order to form the fitted super-resolution edge coordinate set.

[0033] Furthermore, the real-time status parameters of the edge banding equipment are acquired, and the status parameters are normalized to obtain correction factors for correcting measurement deviations, including:

[0034] Obtain the current status parameters of the edge banding equipment, including roller wear, temperature, pressure, motor speed, and sensor calibration error.

[0035] Each state parameter is normalized to obtain the normalized state parameters;

[0036] Based on the normalized state parameters, determine the center value and sensitivity of the Gaussian function corresponding to each state parameter; based on the center value and sensitivity of the Gaussian function corresponding to each state parameter, calculate the Gaussian weight of each state parameter, and normalize the Gaussian weight of each state parameter to obtain the normalized Gaussian weight of each state parameter.

[0037] For each state parameter, calculate the corresponding mean and standard deviation from historical data;

[0038] Calculate the coefficient of variation based on the mean and standard deviation of each state parameter, and use the coefficient of variation to calculate the initial weight of each state parameter.

[0039] The normalized Gaussian weight values ​​are multiplied by the corresponding initial weights to obtain the final weights for each state parameter.

[0040] The correction factor is obtained by multiplying each normalized state parameter by its corresponding final weight and summing the results.

[0041] Furthermore, the fitted super-resolution edge coordinate set is corrected using a correction factor to obtain a corrected edge coordinate set, including:

[0042] Iterate through each point in the fitted coordinate set, and multiply the horizontal coordinate (x) and vertical coordinate (y) of each point by a correction factor to calculate the corrected coordinates.

[0043] Furthermore, based on the corrected edge coordinate set, the actual geometric dimensions of the board and the deviation of the actual geometric dimensions are calculated, and a dimensional deviation score is obtained, including:

[0044] Calculate the curvature between adjacent points on the corrected coordinate set, identify curvature abrupt change points, and divide the coordinate set into straight line segments and circular arc segments according to the curvature threshold;

[0045] For straight segments, the lengths of all straight segments are summed to obtain the total length and width of the board; for arc segments, the average value of the fillet radius is taken as the overall fillet feature.

[0046] Calculate the length and width deviations based on the total length and width of the board; calculate the corner radius deviations based on the overall corner radius characteristics.

[0047] Calculate the corresponding relative deviations of length, width, and fillet radius based on the length deviation, width deviation, and fillet radius deviation, respectively.

[0048] The relative deviations in length, width, and fillet radius are normalized, and the weighted sum of the normalized deviations is calculated to generate a dimensional deviation score.

[0049] Furthermore, based on the dimensional deviation score, boards that do not meet the requirements are marked, and the equipment is controlled to perform corresponding processing, including:

[0050] Set a scoring threshold. If the board score is greater than the scoring threshold, it is judged as unqualified, and a mark, deviation type and score value are added to the board surface.

[0051] The marked material is transferred to the rework station, where preset cutting, grinding, or shaping parameters are called.

[0052] Secondly, an intelligent control system for a fully automatic edge banding equipment for sheet metal includes:

[0053] The determination module is used to construct an octree structure and map pixel values ​​based on the original board image resolution and preset octree depth. The nodes are decomposed into subband sets through multiple filters. The high-frequency subbands are morphologically processed to obtain a binarized edge map. After determining the edge points, Hough transform is used to connect them into curves and interpolation is used to extract the super-resolution edge coordinates.

[0054] The calculation module is used to perform median filtering on the super-resolution edge coordinates and sort them clockwise to form an ordered set. It calculates the curvature markers to divide the edge segments, randomly selects points for each segment as a unit to fit a circle and optimizes it to obtain the final arc model. The sampling points along the arc model are mapped back to the original image coordinate system and merged to form the fitted super-resolution edge coordinate set.

[0055] The normalization module is used to acquire the real-time status parameters of the edge banding equipment, and to normalize the status parameters to obtain a correction factor for correcting measurement deviations.

[0056] The correction module is used to correct the fitted super-resolution edge coordinate set by a correction factor to obtain the corrected edge coordinate set.

[0057] The evaluation module is used to calculate the actual geometric dimensions of the board and the deviation of the actual geometric dimensions based on the corrected edge coordinate set, and to obtain the dimensional deviation score.

[0058] The control module is used to mark boards that do not meet the requirements based on dimensional deviation scores and control the equipment to perform corresponding processing.

[0059] Thirdly, a computing device, comprising:

[0060] One or more processors;

[0061] A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0062] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0063] The above-described solutions of the present invention include at least the following beneficial effects.

[0064] By constructing an octree structure and mapping pixel values, combined with multi-layer decomposition of filters and morphological processing, this invention can accurately extract the binarized image of the edge of the board material. Then, by using Hough transform to connect the edge points and interpolation to extract the super-resolution edge coordinates, this process significantly improves the accuracy of edge recognition.

[0065] After median filtering and sorting of the super-resolution edge coordinates, the edge segments are divided by calculating the curvature markers and fitting the arc model with each segment as a unit. This step not only optimizes the set of edge coordinates, but also makes the edge fitting more in line with the actual shape, improving the fit and aesthetics of the edge sealing.

[0066] By acquiring and normalizing the real-time status parameters of the edge banding equipment, a correction factor is obtained to correct measurement deviations. This invention enables the edge banding operation to be adjusted in real time to compensate for errors caused by changes in equipment status, thereby improving the accuracy and stability of edge banding.

[0067] By correcting the fitted super-resolution edge coordinate set with a correction factor, a more accurate edge coordinate set is obtained. Based on these coordinate sets, the present invention can accurately calculate the actual geometric dimensions of the board and its deviations.

[0068] Based on the dimensional deviation score, the present invention can automatically mark the boards that do not meet the requirements and control the equipment to perform corresponding processing. This intelligent processing flow not only improves production efficiency, but also reduces the cost and error of manual intervention, making the edge banding process more automated and intelligent.

[0069] In summary, this invention significantly improves the quality and production efficiency of board edge banding through a series of steps, including high-precision edge recognition, edge fitting and coordinate optimization, real-time state parameter correction, edge coordinate correction and size calculation, and size deviation scoring and intelligent processing. At the same time, this method also reduces the skill requirements of operators and reduces the impact of human factors on product quality. Attached Figure Description

[0070] Figure 1 This is a flowchart illustrating an intelligent control method for a fully automatic edge banding equipment for sheet metal, provided by an embodiment of the present invention. Detailed Implementation

[0071] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0072] like Figure 1 As shown in the figure, an embodiment of the present invention proposes an intelligent control method for a fully automatic edge banding equipment for sheet metal, the method comprising the following steps:

[0073] Step 1: Construct an octree structure and map pixel values ​​based on the original board image resolution and preset octree depth. Decompose the nodes through multiple layers of filters to obtain subband sets. Perform morphological processing on the high-frequency subbands to obtain a binarized edge map. After determining the edge points, connect them into curves using Hough transform and extract the super-resolution edge coordinates by interpolation.

[0074] Step 2: After median filtering of the super-resolution edge coordinates, sort them clockwise to form an ordered set. Calculate the curvature markers to divide the edge segments. Use each segment as a unit to randomly select points to fit a circle and optimize it to obtain the final arc model. Map the sampling points along the arc model back to the original image coordinate system and merge them to form the fitted super-resolution edge coordinate set.

[0075] Step 3: Obtain the real-time status parameters of the edge banding equipment, and normalize the status parameters to obtain the correction factor used to correct the measurement deviation.

[0076] Step 4: Correct the fitted super-resolution edge coordinate set using a correction factor to obtain the corrected edge coordinate set;

[0077] Step 5: Based on the corrected edge coordinate set, calculate the actual geometric dimensions of the board and the deviation of the actual geometric dimensions, and obtain the dimensional deviation score;

[0078] Step 6: Based on the dimensional deviation score, mark the boards that do not meet the requirements and control the equipment to perform the corresponding processing.

[0079] In this embodiment of the invention, step 1, by constructing an octree structure based on the original board image resolution and a preset octree depth, can efficiently organize and store image data. The mapped pixel values ​​retain detailed information in the image, providing a rich data source for edge recognition. Multi-level filter decomposition of nodes can extract different frequency components in the image, especially the high-frequency subband, which contains rich edge information. This step helps to identify edges more accurately in subsequent steps. Morphological processing of the high-frequency subband can highlight edge features and suppress noise, resulting in a clear binarized edge map. This greatly simplifies the complexity of edge recognition and improves its accuracy. Connecting edge points into curves using Hough transform and interpolating to extract super-resolution edge coordinates allows for precise edge location.

[0080] Step 2 involves median filtering of the super-resolution edge coordinates, effectively removing noise points and smoothing the edges. Sorting the coordinates clockwise creates an ordered set, facilitating subsequent edge fitting and processing. Calculating curvature and marking segmentation points accurately divides the edge segments, helping to better adapt to edges of different shapes and improve the sealing fit. Randomly selecting points on each segment to fit a circle, and optimizing it to obtain the final arc model, more accurately describes the edge shape. Mapping the sampled points along the arc model back to the original coordinate system and merging them forms the fitted super-resolution edge coordinate set, providing more precise data for the sealing operation. Obtaining the real-time status parameters of the sealing equipment allows for timely understanding of its operating status and performance changes. Normalizing the status parameters eliminates dimensional differences between different parameters, facilitating comparison and analysis. The calculated correction factor can be used to correct measurement deviations and improve the accuracy of the sealing process.

[0081] Step 4: Correcting the fitted super-resolution edge coordinate set using a correction factor eliminates errors caused by changes in equipment status, resulting in a more accurate edge coordinate set. This step helps improve the precision and stability of edge banding. Step 5: Based on the corrected edge coordinate set, the actual geometric dimensions of the board can be accurately calculated, along with the deviations, resulting in a dimensional deviation score. This score directly reflects the dimensional accuracy and edge banding quality, facilitating the timely detection and handling of defective products. Step 6: Marking boards that do not meet the requirements based on the dimensional deviation score. This step improves production efficiency and product quality, allowing the equipment to handle issues appropriately, such as adjusting edge banding parameters or changing edge banding materials. It automatically adapts to the edge banding needs of different boards, increasing the flexibility and adaptability of edge banding. Simultaneously, it reduces the skill requirements of operators and minimizes the impact of human factors on product quality.

[0082] In a preferred embodiment of the present invention, an octagonal structure is constructed based on the original plate image resolution and a preset octagonal depth, and pixel values ​​are mapped. The nodes are then decomposed through multiple layers of filters to obtain a set of sub-bands, including:

[0083] Based on the resolution of the original board image and the preset octree depth, an octree structure is constructed. Each node in the octree structure represents a pixel block in the original board image. Specifically, this includes: determining the resolution of the original board image, for example, width W and height H; and calculating the size of the pixel block represented by each node based on the preset octree depth D, assuming the image is uniformly divided into 2... D ×2 D In a grid, the size of each node represents a pixel block of size. Initialize an octree structure, where each node contains its location information (such as coordinates), pixel block size, and pointers to child nodes. Recursively construct the octree, starting from the root node, dividing the image into smaller sub-blocks according to the pixel block size, and creating a child node for each sub-block, until the preset octree depth is reached.

[0084] The pixel values ​​of the original board image are mapped block by block to the nodes of the octree structure. Specifically, this includes: traversing each pixel of the original board image and calculating which octree node it belongs to based on its coordinates; accumulating the pixel value of the pixel into the sum of the pixel values ​​of the corresponding node; if the node represents a pixel block, the values ​​of all pixels in the pixel block can be stored.

[0085] For each node of the octree, convolution operations are performed using low-pass and high-pass filters of Haar wavelets. After each decomposition, one low-frequency subband and three high-frequency subbands are generated. Specifically, for each node in the octree (representing a pixel block), its pixel value is obtained (if the node stores the values ​​of all pixels within the pixel block, it is obtained directly; if it stores the feature representation, the pixel block needs to be reconstructed first). The pixel block is then convolved using low-pass and high-pass filters of Haar wavelets. The Haar wavelet filter is a 2×2 matrix. The low-pass filter is used to smooth the image, and the high-pass filter is used to detect edges. After the convolution operation, four subbands are obtained: one low-frequency subband (LL) and three high-frequency subbands (LH, HL, HH), representing different frequency components of the image. The four subbands obtained from the convolution operation are stored separately. The low-frequency subband (LL) will be used for the decomposition of the next layer, while the high-frequency subbands (LH, HL, HH) are retained as high-frequency information of the current layer.

[0086] The low-frequency subbands obtained from the current layer decomposition are used as input for the next layer decomposition. The low-frequency subband data is redistributed to the nodes of the octree, and the node size is adjusted to match the resolution after the current decomposition. The filtering operation is repeated on the remapped low-frequency subbands to generate new low-frequency and high-frequency subbands. If the current layer number has not reached the preset layer number, the layer number is incremented and the filtering operation is repeated; if the preset layer number has been reached, the decomposition is terminated to obtain a set of subbands after multi-layer decomposition. The subband set includes the low-frequency subbands of the final layer and the high-frequency subbands of each layer, specifically including:

[0087] Pixel values ​​are extracted from the low-frequency subband (LL) obtained from the current layer decomposition and used as input for the next layer decomposition, i.e., as pixel values ​​for the octree nodes in the next layer. The size of the octree nodes is adjusted according to the resolution after the current decomposition. Since each layer decomposition halves the resolution of the image (in both the horizontal and vertical directions), the size of the nodes also needs to be halved accordingly. The pixel values ​​of the low-frequency subband are redistributed to the adjusted octree nodes. This requires rebuilding part or all of the octree to adapt to the new resolution. The remapped low-frequency subband is then convolved again with a low-pass filter and a high-pass filter of Haar wavelet. This process is repeated to generate a new low-frequency subband and three high-frequency subbands.

[0088] A counter for the current layer number is maintained. After each layer decomposition is completed, it is checked whether the current layer number has reached the preset layer number. If not, the layer counter is incremented and the decomposition continues. If the preset layer number has been reached, the decomposition process is terminated. During the decomposition process, the high-frequency subbands (LH, HL, HH) generated by each layer are saved. When the decomposition terminates, the low-frequency subband (LL) of the final layer is also saved. All the saved subbands are combined into a subband set as input for subsequent processing. This process achieves efficient representation and feature extraction of the original board image through steps such as constructing an octree structure, mapping pixel values, and multi-layer filter decomposition.

[0089] In this embodiment of the invention, the octree structure can efficiently divide a three-dimensional (or, in this scenario, a hierarchical division of a two-dimensional image) space. Each node represents a pixel block in the original board image. This division method helps to quickly locate and access specific regions in the image, improving the efficiency of subsequent processing. By setting a preset octree depth, image representations at different scales can be constructed, which helps to consider both local details and global features of the image in subsequent processing, improving the accuracy of edge recognition. Mapping pixel values ​​to octree nodes allows for the orderly organization of image data, which helps to quickly access and modify pixel values ​​in specific regions in subsequent processing, improving processing efficiency. Block-by-block mapping preserves the detailed information in the image, providing a rich data source for subsequent edge recognition and super-resolution reconstruction. Through convolution operations of low-pass and high-pass filters, the image can be decomposed into components of different frequencies. The low-frequency subband contains the main information and contours of the image, while the high-frequency subband contains the details and edge information of the image. The high-pass filter helps to highlight the edge features in the image, making subsequent edge recognition more accurate.

[0090] Each decomposition layer generates one low-frequency subband and three high-frequency subbands, providing rich image information. By adjusting the number of decomposition layers, the richness of subband information and the degree of detail preservation can be flexibly controlled, which helps to customize processing according to the needs of different tasks. By decomposing the low-frequency subband layer by layer, the feature representation of the image can be gradually refined, which helps to more accurately identify edge and detail information in subsequent processing. Since the low-frequency subband contains the main information of the image, decomposing it layer by layer can reduce the amount of computation and improve processing efficiency. As the number of decomposition layers increases, the image resolution gradually decreases. By adjusting the size of the octree nodes to match the resolution after the current decomposition, the effective representation of image data in the octree structure can be ensured. The data of the low-frequency subband is redistributed to the octree nodes, maintaining data consistency and integrity. By repeating the filtering operation, the feature representation of the image can be continuously refined, improving the accuracy of edge recognition. Each decomposition layer generates new low-frequency and high-frequency subbands, providing multi-layer feature information.

[0091] By controlling the decomposition process by setting a preset number of layers, it is possible to ensure that the required subband information is obtained. When the preset number of layers is reached, the decomposition is terminated, which can avoid the increase in computation and information loss caused by over-decomposition. This helps to maintain a balance between processing efficiency and accuracy. The subband set after multi-layer decomposition provides rich image feature representations, including information such as contours, details and edges.

[0092] In a preferred embodiment of the present invention, a binarized edge map is obtained by morphological processing of the high-frequency subband. After determining the edge points, a curve is connected by Hough transform and the super-resolution edge coordinates are extracted by interpolation, including:

[0093] Morphological processing is performed on the high-frequency subband. In the morphologically processed high-frequency subband, edge points are determined through thresholding. Specifically, this includes: processing the high-frequency subband using erosion operations, for example, using structuring elements (such as 3×3 rectangles or crosses) to remove small noise points and highlight edge features; and performing thresholding on the morphologically processed high-frequency subband, selecting a suitable threshold, setting regions with pixel values ​​above the threshold to white (or 1), and regions below the threshold to black (or 0). Alternatively, the threshold can be automatically calculated using the Otsu algorithm. In the binarized image, white regions represent edge points, which can be identified by traversing the image pixels and recording the coordinates of pixels with a value of 1 as edge points.

[0094] The process involves inputting edge points into the Hough transform algorithm, identifying collinear points and fitting them into a continuous curve, and merging adjacent curve segments to form a continuous edge curve. Specifically, this includes: inputting the identified edge points into the Hough transform algorithm, which can detect straight lines in the image; recording the number of points on each possible straight line through an accumulator array; selecting the straight line with the higher value in the accumulator array as a candidate straight line; identifying the edge points (collinear points) on each candidate straight line; using these collinear points for curve fitting, such as fitting a straight line or a polynomial curve using the least squares method; and checking the connectivity and directional consistency between adjacent curve segments. If the angle between adjacent curve segments is less than a certain threshold and they are spatially close enough, they are merged into a continuous curve.

[0095] On a continuous edge curve, linear interpolation is used to obtain the coordinates of the interpolated edge points. Specifically, this involves selecting points to be interpolated on the continuous edge curve. These points can be any location on the curve, but typically connection points between curve segments or key points on the curve are chosen for interpolation. For each selected interpolation point, linear interpolation is performed using its two adjacent known points. The formula for linear interpolation is:

[0096]

[0097] Here, (x1, y1) and (x2, y2) are two adjacent known points, and (x, y) are the coordinates of the interpolation point. The coordinates of the interpolated edge point can be obtained through linear interpolation.

[0098] Extracting the coordinates of interpolated edge points to form a super-resolution edge coordinate set involves: traversing the interpolated edge points and extracting their coordinates, which can be represented as (x, y), where x and y are the horizontal and vertical coordinates of the edge point in the image, respectively. Combining all the extracted edge point coordinates into a set, i.e., the super-resolution edge coordinate set.

[0099] In this embodiment of the invention, morphological processing (such as dilation, erosion, opening operation, closing operation, etc.) can effectively highlight the edge features in the high-frequency subband, reduce noise interference, and make the edges clearer. Through threshold segmentation, the morphologically processed high-frequency subband can be converted into a binary edge map, making the determination of edge points simpler and more accurate. The combination of morphological processing and threshold segmentation can significantly improve the accuracy of edge detection. The Hough transform can identify collinear points among edge points, thus fitting them into a continuous curve. Through the Hough transform, discrete edge points can be connected into a continuous curve, better describing the shape and direction of the edge. Adjacent curve segments can be merged using certain algorithms to form a more complete continuous edge curve, improving the coherence and accuracy of the edge. Linear interpolation can generate more edge point coordinates on the continuous edge curve, thereby improving the edge resolution and making the edges smoother and more delicate. The interpolated edge point coordinates can better capture the detailed information of the edge, and the interpolated edge point coordinates are denser and more uniform. Super-resolution edge coordinate sets contain richer edge information and can be used to improve image resolution and quality. Image reconstruction or edge rendering using super-resolution edge coordinate sets can achieve clearer and more delicate edge effects, enhancing the visual effect of the image.

[0100] In a preferred embodiment of the present invention, after median filtering of the super-resolution edge coordinates, they are sorted clockwise to form an ordered set, and curvature marker segmentation points are calculated to divide the edge segments, including:

[0101] Median filtering is applied to the super-resolution edge coordinates. The filtered coordinates are sorted clockwise along the edge curve to form an ordered set of edge coordinates. Specifically, this involves: determining the window size for median filtering, typically an odd number such as 3×3 or 5×5, the window size depending on the density of the edge coordinates and the noise level; for each edge coordinate point, selecting all coordinate points (including the point itself) within a filtering window centered on that point; sorting the x and y coordinates of all points within the filtering window, then taking the median as the filtered coordinate for that point; and applying the median filter... The original coordinates are replaced by the filtered coordinates to form a set of edge coordinates. A suitable starting point is selected, such as the point with the smallest x-coordinate or the smallest y-coordinate in the edge coordinate set, as the starting point for sorting. For each edge coordinate point, the angle (polar angle) between the line connecting it to the starting point and the positive x-axis is calculated. The arctangent function (such as atan2) can be used to calculate the polar angle. All edge coordinate points are sorted according to the size of the polar angle. If the polar angles are the same, they can be sorted a second time according to their distance from the starting point. The sorted edge coordinate points are then combined into an ordered set of edge coordinates.

[0102] The local curvature of ordered coordinate points is calculated using cubic spline interpolation. This process iterates through all points, marking any point whose curvature exceeds a threshold as a segmentation point. Specifically, this involves: constructing a cubic spline interpolation function using the ordered edge coordinate set. This function fits a smooth curve that passes through all ordered edge coordinate points. The first and second derivatives of the cubic spline interpolation function are calculated, where the first derivative represents the slope of the curve and the second derivative represents the rate of change of curvature. Based on the first and second derivatives, the local curvature at each coordinate point is calculated. An appropriate curvature threshold is empirically set; when the curvature exceeds this threshold, the point is considered a segmentation point. This process is repeated for all ordered edge coordinate points, calculating the curvature at each point. If the curvature of a point exceeds the threshold, that point is marked as a segmentation point. Specific applications are as follows:

[0103] Let the ordered set of edge coordinate points be {(x0,y0),(x1,y1),…,(x n ,y n Arrange them in clockwise or counterclockwise order; parameterize the curve as t∈[0,n], where n represents the number of edge coordinate points, and t=i corresponds to the i-th point (x i ,y i );

[0104] Construct cubic spline functions S for x(t) and y(t) respectively. x (t) and S y (t), such that:

[0105] S x (t i )=x i ,S y (t i )=y i (i = 0, 1, ..., n);

[0106] Interpolation methods:

[0107] Each interval [t] i ,t i+1 Within [the definition], a cubic polynomial is defined as follows:

[0108] S(t) = a i +b i (tt i )+c i (tt i ) 2 +d i (tt i ) 3 ;

[0109] Satisfy the continuity condition:

[0110] Positional continuity: S(t)i )=y i ;

[0111] The first derivative is continuous:

[0112] The second derivative is continuous:

[0113] Boundary conditions (such as natural splines):

[0114] S″(t0)=S″(t n ) = 0;

[0115] Calculate the first and second derivatives:

[0116] With S x (t) and S y For example, (t):

[0117] First derivative:

[0118] S′ x (t)=b i +2c i (tt i )+3d i (tt i ) 2 ;

[0119] S′ y (t)=b i +2c i (tt i )+3d i (tt i ) 2 ;

[0120] Second derivative:

[0121] S″ x (t)=2c i +6d i (tt i );

[0122] S″ y (t)=2c i +6d i (tt i );

[0123] The local curvature is calculated, where the formula for curvature κ(t) is:

[0124]

[0125] The threshold κ is determined through statistical analysis based on the specific application scenario. h(For example: take the top 5 percentiles of the curvature distribution, or manually adjust to achieve the best segmentation effect);

[0126] Traverse all points and mark the split points. The traversal process is as follows:

[0127] For each point (x) i ,y i Calculate the curvature κ at the corresponding parameter t = i. i If κ i >κ h , mark (x i ,y i Let be the dividing point, where the set of dividing points is .

[0128] Where x(t) represents the x-coordinate of the ordered edge point as a function of parameter t; y(t) represents the y-coordinate of the ordered edge point as a function of parameter t; t i This represents the parameter value corresponding to the i-th point. If index parameterization is used, t i =i (e.g., t0 = 0, t1 = 1); t i+1 This represents the parameter value corresponding to the (i+1)th point, in the range [t]. i ,t i+1 ];a i Represents the interval [t] i ,t i+1 The constant term of the cubic polynomial is equal to S(t). i The value of ) (i.e., x) i or y i );b i Let t = t i The first derivative value at c; i Denotes the coefficient of the quadratic term, and represents t = t i Half the value of the second derivative at that point (since the second derivative is 2c) i +6d i (tt i ));d i The coefficient of the cubic term is denoted as , which controls the curvature of the curve within the interval; t represents the parameterized variable, whose value range covers all ordered points, for example, t∈[0,n]; S′ x (t) represents the first derivative of x(t) with respect to the parameter t, and represents the rate of change in the x-direction; S′ y (t) represents the first derivative of y(t) with respect to the parameter t, and represents the rate of change in the y-direction; S″ x (t) represents the second derivative of x(t) with respect to the parameter t, and represents the rate of change of curvature in the x-direction; S″ y (t) represents the second derivative of y(t) with respect to parameter t, and represents the rate of change of curvature in the y direction.

[0129] The edge is divided into multiple continuous segments based on the split points. Specifically, this involves: creating an empty segment list to store the divided edge segments; starting from the starting point of the ordered edge coordinate set, traversing all coordinate points; when a split point is encountered, adding the current segment (all coordinate points from the previous split point to the current split point) to the segment list and starting a new segment; after the traversal is complete, adding the last segment (all coordinate points from the last split point to the end point) to the segment list. Each element in the segment list is a continuous edge segment, and these segments together constitute the divided edge segment set.

[0130] In this embodiment of the invention, median filtering is an effective nonlinear filtering method that can remove noise points in the edge coordinates and preserve the true structure of the edge. This is crucial for subsequent edge analysis and processing because noise points may interfere with curvature calculation and accurate marking of segmentation points. Median filtering can also smooth the edges to a certain extent, making them more continuous and smooth, which helps improve the accuracy and stability of subsequent curvature calculations. An ordered set of edge coordinates makes subsequent processing (such as curvature calculation and segmentation point marking) more convenient and efficient because an ordered set of coordinates ensures that processing is performed according to the actual direction of the edge, avoiding processing errors caused by disordered coordinate points. Clockwise sorting preserves the topological structure of the edge, i.e., the starting and ending points of the edge and the connection relationships between them.

[0131] Cubic spline interpolation can accurately fit edge curves and calculate local curvature. It can capture subtle changes in edges, including abrupt and gradual changes in curvature, which helps to more accurately identify edge segmentation points. By setting a curvature threshold, edge segmentation points can be automatically marked, dividing the edge into multiple continuous segments. This helps to more clearly understand the edge's structure and features. The curvature threshold can be adjusted according to the actual situation to ensure the accuracy and stability of segmentation. By appropriately selecting the threshold, over-segmentation or under-segmentation can be avoided. After dividing the edge into multiple continuous segments, each segment can be processed and analyzed independently, such as feature extraction and shape recognition. Segmented processing can significantly improve processing efficiency.

[0132] In a preferred embodiment of the present invention, points are randomly selected as units to fit a circle and optimized to obtain the final arc model. The sampled points along the arc model are mapped back to the original image coordinate system and merged to form a fitted super-resolution edge coordinate set, including:

[0133] Extracting the set of coordinate points within each segment as an independent fitting unit involves: obtaining the edge segment information from the previous steps, where each segment contains a set of continuous coordinate points; extracting all coordinate points within each segment to form an independent set of coordinate points; and storing the set of coordinate points for each segment as an independent fitting unit.

[0134] Based on the set of coordinate points for each segment, three points are randomly selected from the coordinate points of the current segment. The center and radius of the fitted circle are calculated. For all points within the segment, the distance to the current arc model is calculated. Specifically, for each fitting unit (i.e., the set of coordinate points for the segment), three non-collinear points are randomly selected. Using these three points, the center and radius of the fitted circle are calculated using geometric methods (such as the perpendicular bisector method). For all points within the fitting unit, their distance to the center of the current fitted circle is calculated and compared with the radius of the circle to obtain the distance from each point to the arc model. In practical applications, this may include:

[0135] Three points are randomly selected from the set of coordinate points of the fitted unit, denoted as A(x1,y1), B(x2,y2), and C(x3,y3); these three points are ensured to be non-collinear, which can be achieved by calculating vectors. and We can use the cross product to determine if the three points are collinear. If the cross product is not zero, then the three points are not collinear.

[0136] Calculate the equations of the perpendicular bisectors of line segments AB and AC:

[0137] The coordinates of the midpoint of line segment AB:

[0138] The slope of line segment AB: (when x1≠x2);

[0139] The slope of the perpendicular bisector of line segment AB (when k) AB ≠0); if k AB =0, then the perpendicular bisector is a vertical line.

[0140] The equation of the perpendicular bisector of line segment AB: using the point-slope form. Where y and x represent the coordinate variables of any point on the perpendicular bisector; and M represents the midpoint of line segment AB. AB The coordinates, where: (The average of the x-coordinates of the two endpoints of line segment AB); (The average of the ordinates of the two endpoints of line segment AB).

[0141] Similarly, calculate the equation of the perpendicular bisector of line segment AC.

[0142] Solve the equations of the perpendicular bisectors of line segments AB and AC, and obtain the coordinates of the intersection point (x0, y0), which is the center of the fitted circle.

[0143] Calculate the radius using the center (x0, y0) and any point (such as point A).

[0144] For each point P(x) within the fitting unit i ,y i ), calculate its distance to the center (x0, y0):

[0145]

[0146] The calculated distance d i By comparing the distance from each point to the circular arc model with the radius r of the circle, we obtain the distance difference (or residual) Δd. i =|d i -r|.

[0147] Points with a distance less than a threshold are marked as inliers. The center and radius of the circle are then re-optimized using the least squares method with all inliers to obtain the final arc model for each segment. Specifically, this involves: setting an appropriate distance threshold based on experience; marking points with a distance less than the threshold as inliers (these points are considered to fit the current arc model); and refitting the center and radius of the circle using the least squares method with all inliers to obtain a more accurate arc model. The least squares method minimizes the sum of squared distances from inliers to the arc model, thus obtaining the optimal center and radius. In practical applications, this may include:

[0148] Based on experience, a suitable distance threshold is set, which is used to determine whether a point conforms to the current circular arc model; for each point P within the fitting unit... i (x i ,y i ), calculate the distance from the current fitted circle center (x0, y0). If d i If the distance is less than or equal to the inner circle, then mark the point as an interior point; minimize the sum of squared distances from the interior point to the circular arc model, which is to solve the following optimization problem:

[0149]

[0150] Among them, (x c ,y c ) is the new center coordinate, and r is the new radius.

[0151] Algebraic fitting methods (based on the general equation of a circle):

[0152] The general equation of a circle:

[0153] x 2 +y 2 +Dx+Ey+F=0;

[0154] Wherein, the coordinates of the center are radius is

[0155] For each interior point (x) i ,y i Substituting this into the general equation of a circle, we get:

[0156]

[0157] This can be viewed as a linear equation concerning D, E, and F.

[0158] Combine the equations of all interior points into a system of linear equations, and solve this system of linear equations using the least squares method (such as matrix solving or iterative optimization methods) to obtain the optimal solutions for D, E, and F.

[0159] Calculate the center and radius of the circle:

[0160] Center coordinates:

[0161] radius:

[0162] For the final circular arc model of each segment, points are sampled uniformly along the circular arc path with a super-resolution step size, specifically including:

[0163] For circles or arcs, use parametric equations to generate sampling points:

[0164]

[0165] Where (x0, y0) are the coordinates of the center of the circle, r is the radius, and θ is the parameter angle. The angle step size Δθ is set according to the super-resolution requirements. The smaller the step size, the denser the sampling points. The total number of sampling points N is determined by the arc span Δθ. total Determined by step size:

[0166]

[0167] From the starting angle θ start Initially, the step size Δθ is incremented sequentially to generate the parameter sequence:

[0168] θ i =θ start +i·Δθ(i=0,1,2,…,N-1);

[0169] Substitute the parameters into the equation and calculate the coordinates of each sampling point:

[0170] P i =(x0+rcos(θ) i ),y0+rsin(θ i ));

[0171] Where, θ i It is the parameter angle of the i-th sampling point, which is obtained from the starting angle θ. start Initially, it is generated by incrementing by a fixed angular step size Δθ each time, where θ i The position of the sampling point on the arc was determined; θ start θ is the parameter angle at which sampling begins. It typically corresponds to the angle between the starting point of an arc or circle and the center of the circle. In practical applications, θ start The step size can be determined based on the starting position and direction of the arc; Δθ is the angular step size, which is the angular difference between adjacent sampling points. It determines the density of the sampling points. The smaller the step size, the denser the sampling points, and the better the super-resolution effect; i is the index of the sampling point, which starts from 0 and increments. It is used to locate each sampling point in the parameter sequence; N is the total number of sampling points, which is determined by the total span of the arc (in degrees) and the angular step size Δθ. For example, if the total span of the arc is 90° and Δθ = 1°, then N = 90; P i `x` represents the coordinates of the i-th sampling point, a two-dimensional point composed of x and y coordinates; `x0` and `y0` are the x and y coordinates of the circle's center, defining the position of the circle or arc on the two-dimensional plane; `r` is the radius of the circle, determining its size; `cos(θ)`... i ) and sin(θ i These two are trigonometric functions, used to calculate the parametric angle θ. i The cosine and sine values.

[0172] The sampling points are mapped back to the original image coordinate system, and the dense coordinate points of all segments are merged according to the original sorting order to form the fitted super-resolution edge coordinate set, which specifically includes:

[0173] Coordinate system definition:

[0174] Local coordinate system: The origin is the center of the circle (x0, y0), and the X-axis is along the direction from the center of the circle to the starting point.

[0175] Original image coordinate system: with the top left corner of the image as the origin, the X-axis to the right and the Y-axis downward.

[0176] The origin (x0, y0) of the local coordinate system is translated to the corresponding position in the image coordinate system. The translation vector is:

[0177] T = (x0, y0);

[0178] Coordinates after translation:

[0179] P′ i =P i +T=(x0+rcos(θ i ),y0+rsin(θ i ));

[0180] Rotation transformation (if the local coordinate system is rotated relative to the image coordinate system):

[0181] The rotation angle is calculated to determine the angle α between the local coordinate system's X-axis and the image coordinate system's X-axis. Typically, α can be determined from the starting point P. start Calculate with the center (x0, y0):

[0182] α=arctan2(y start -y0,x start -x0);

[0183] Rotation matrix:

[0184]

[0185] Coordinates after rotation:

[0186] P″ i =R(α)·(P i -T)+T;

[0187]

[0188] Where, x″ i and y″ i These two are the x and y coordinates of the i-th sampled point after transformation in the original image coordinate system, respectively; the two parameters x0 and y0 are the x and y coordinates of the center of the circle, respectively, defined in the original image coordinate system;

[0189] For each segment, record its starting index k in the original edge point set. start and the ending index k end Initialize an empty list P merged ; Traverse all segments in the original order, and sample the points {P″ of each segment. i}Append to P sequentially merged In the middle; since the sampling points have been generated in increments according to the parameter angle θ, and the merging is performed in the order of the original segmentation, the merged point set naturally retains the topological order of the original edge points.

[0190] In this embodiment of the invention, the complex edge curve is divided into multiple simple segments, each of which can be fitted independently, reducing the difficulty and complexity of fitting. The set of coordinate points within each segment retains the local features of that segment, helping to fit the arc model of that segment more accurately. Randomly selecting three points for initial fitting increases the diversity and robustness of the fitting, helping to avoid local optima. The distance from all points within the segment to the current arc model is calculated. Marking interior points with a distance threshold helps to remove noise and outliers, improving the accuracy of fitting. Using all interior points, the center and radius are re-optimized using the least squares method, resulting in a more accurate arc model and better fitting of the coordinate points within the segment. Uniformly sampling points with a super-resolution step size generates more points on the arc, improving the resolution and detail of the edges. The even distribution of sampling points on the arc helps to more accurately describe the shape and features of the arc. The sampling points are mapped back to the original image coordinate system, so that the fitted edge coordinate set is consistent with the original image, which facilitates subsequent processing and analysis. The dense coordinate points of all segments are combined in the original sorting order to form a complete fitted super-resolution edge coordinate set, which preserves the integrity and continuity of the edges.

[0191] In a preferred embodiment of the present invention, real-time status parameters of the edge-sealing equipment are obtained, and the status parameters are normalized to obtain a correction factor for correcting measurement deviations, including:

[0192] The system acquires the current status parameters of the edge banding equipment, including roller wear, temperature, pressure, motor speed, and sensor calibration error values. Specifically, it collects the equipment's status parameters in real time using various sensors installed on the edge banding equipment (such as wear sensors, temperature sensors, pressure sensors, speed sensors, and calibration error sensors). The collected data is then transmitted to a data processing unit (such as a PLC, computer, or embedded system) via wired or wireless means. The data processing unit receives the sensor data and stores it in memory or a database for subsequent processing.

[0193] Normalization is performed on each state parameter to obtain normalized state parameters. Specifically, this includes: for each state parameter, determining its normalization range, for example, [0, 1], and using... Map the state parameters to a normalized range, where x is the original state parameter value. min and x max These are the minimum and maximum values ​​of the parameter (which can be determined based on equipment specifications or historical data). Applying the above formula to each state parameter yields the normalized state parameter x. norm .

[0194] Based on the normalized state parameters, determine the center value and sensitivity of the Gaussian function corresponding to each state parameter; calculate the Gaussian weight of each state parameter based on the center value and sensitivity of the Gaussian function corresponding to each state parameter, and normalize the Gaussian weight of each state parameter to obtain the normalized Gaussian weight of each state parameter. Specifically, this includes: determining the center value of the Gaussian function corresponding to each parameter based on the normalized state parameters, setting the center value as the average value of the parameter within its normal operating range; the sensitivity reflects the degree of influence of parameter changes on the correction factor, and the sensitivity value of each parameter can be determined experimentally or empirically; and calculating the Gaussian weight w of each state parameter using the Gaussian function formula. i :

[0195]

[0196] Where, x norm,i μ is the i-th normalized state parameter. i σ is the center value of the Gaussian function. i It is the standard deviation related to sensitivity; normalize the Gaussian weights of all state parameters so that their sum is 1.

[0197]

[0198] Among them, w norm,i The normalized Gaussian weights are represented by n, and the Gaussian weights w are represented by n. i The quantity.

[0199] For each state parameter, the corresponding mean and standard deviation are calculated from historical data. Specifically, this includes collecting multiple sample values ​​for each state parameter from historical data, calculating the mean and standard deviation of each state parameter.

[0200] Based on the mean and standard deviation of each state parameter, the coefficient of variation is calculated. The initial weight of each state parameter is then calculated using the coefficient of variation. Specifically, this involves calculating the ratio of the standard deviation to the mean to obtain the coefficient of variation for each state parameter, and then using the coefficient of variation to calculate the initial weight of each state parameter. A larger coefficient of variation indicates greater historical volatility of the parameter and a greater impact on the correction factor; therefore, it is assigned a larger initial weight. init,i ,in,

[0201]

[0202] Among them, C Vi Let C represent the i-th coefficient of variation. Vj Let represent the j-th coefficient of variation, and n represent the number of coefficients of variation.

[0203] The normalized Gaussian weight values ​​are multiplied by the corresponding initial weights to obtain the final weights for each state parameter; each normalized state parameter is multiplied by its corresponding final weight and summed to obtain the correction factor.

[0204] In this embodiment of the invention, by acquiring the state parameters of the edge-sealing equipment in real time (such as roller wear, temperature, pressure, motor speed, sensor calibration error, etc.), changes in the equipment state can be detected in a timely manner, thereby more accurately reflecting the actual operating status of the equipment. Normalization of each state parameter eliminates the influence of different dimensions and numerical ranges between parameters, making each parameter comparable in subsequent calculations and improving the accuracy of the correction factor. Based on the normalized state parameters, the center value and sensitivity of the Gaussian function corresponding to each state parameter are determined, and then the Gaussian weight is calculated. The Gaussian function can well describe the degree of influence of the parameter on the correction factor, making the correction factor more consistent with the actual operating characteristics of the equipment. Multiplying the normalized Gaussian weight value by the corresponding initial weight yields the final weight corresponding to each state parameter. This weight adjustment mechanism can dynamically reflect the contribution of each parameter to the correction factor, enhancing the adaptability of the correction factor. The mean and standard deviation of each state parameter are calculated from historical data. The coefficient of variation is used to calculate the initial weight of each parameter, taking into account historical fluctuations. This makes the correction factor more robust; a larger coefficient of variation indicates greater historical fluctuation of the parameter, and thus a greater impact on the correction factor, resulting in a higher initial weight. The correction factor comprehensively considers both the real-time status of the equipment and historical data, enabling more accurate correction of measurement deviations. By applying the correction factor, the operating parameters of the edge banding equipment can be controlled more effectively, reducing the impact of measurement deviations on edge banding quality, thereby improving the quality and consistency of edge banding products. Real-time monitoring and analysis of equipment status parameters can predict potential equipment failures or performance degradation, allowing for proactive maintenance and reducing downtime and repair costs.

[0205] In a preferred embodiment of the present invention, the fitted super-resolution edge coordinate set is corrected by a correction factor to obtain a corrected edge coordinate set, including:

[0206] Iterate through each point in the fitted coordinate set, multiplying the x-coordinate and y-coordinate of each point by a correction factor to calculate the corrected coordinates. Specifically, this includes:

[0207] The fitted super-resolution edge coordinate set is assumed to be Where N is the total number of coordinate points, (x i ,y i ) are the horizontal and vertical coordinates of the i-th point.

[0208] The correction factor is assumed to be CF, used to correct the coordinates; a loop structure is used to iterate through each point in the fitted coordinate set, for each point (x... i ,y i Perform the following calculations:

[0209] Corrected horizontal coordinate: x′ i =x i ×CF;

[0210] Corrected vertical coordinate: y′ i =y i ×CF;

[0211] Where, x′ i and y′ i These are the corrected horizontal and vertical coordinates; the corrected coordinates (x′) i ,y′ i Stored in a new coordinate set, forming the corrected edge coordinate set. Where i represents the index variable, used to traverse each point in the corrected edge coordinate set, with a value range of 1 to N. In the loop or iteration process, i is used to identify the coordinate point currently being processed; N represents the total number of coordinate points in the corrected edge coordinate set.

[0212] In this embodiment of the invention, the fitted super-resolution edge coordinate set may have deviations due to factors such as measurement errors, inaccurate models, or data noise. By multiplying by a correction factor, these errors can be corrected, thereby improving the accuracy of the coordinates. The correction factor can further refine the coordinate values, making the edge coordinate set more accurate at the super-resolution level, which helps to improve the quality of subsequent image processing or analysis.

[0213] The correction factor can be dynamically adjusted based on different device states, measurement conditions, or application scenarios. This means the correction process is more adaptable and can handle various complex situations. By introducing the correction factor, the model can better generalize to unseen data, improving its robustness and stability. Traversing the fitted coordinate set and directly multiplying it by the correction factor is a simple and effective correction method. It does not require complex iterative optimization processes and can quickly obtain the corrected edge coordinate set. This method is easy to implement in existing image processing or analysis workflows without requiring significant modifications to the original algorithm.

[0214] The corrected edge coordinate set can more accurately locate edges in edge detection tasks, improving the accuracy and reliability of edge detection. In image segmentation tasks, the corrected edge coordinate set helps to more accurately divide image regions, improving segmentation results. For applications requiring 3D reconstruction, the corrected edge coordinate set can provide more accurate 3D information, helping to improve reconstruction quality. As the foundational data for subsequent image processing or analysis, the accuracy of the corrected edge coordinate set directly affects the difficulty and effectiveness of subsequent processing. Correcting the coordinates through a correction factor can reduce the difficulty of subsequent processing and improve processing efficiency.

[0215] In a preferred embodiment of the present invention, the actual geometric dimensions of the sheet metal and the deviation of the actual geometric dimensions are calculated based on the corrected edge coordinate set, and a dimensional deviation score is obtained, including:

[0216] Calculate the curvature between adjacent points on the corrected coordinate set, identify curvature abrupt change points, and divide the coordinate set into straight line segments and circular arc segments based on a curvature threshold. Specifically, this includes: for the corrected edge coordinate set... Calculate each point (x′) i ,y′ i The curvature at point () can be approximated by calculating the reciprocal of the radius of the circle formed by three adjacent points, as shown in the following formula:

[0217]

[0218] Among them, κ i Let κ be the curvature at point i, and let κ be a curvature threshold. th When the curvature change between two adjacent points exceeds κ th At that time, the point is considered a point of abrupt change in curvature; based on the point of abrupt change in curvature, the coordinate set is divided into straight line segments and circular arc segments; the straight line segments are continuous point sets with smaller curvature; the circular arc segments are continuous point sets with larger curvature; (x′ i+1 -x′ i (y′) represents the difference between the (i+1)th and ith points in the corrected coordinate set, indicating the change in horizontal distance between adjacent points; i+2 -y′ i+1 (y′) represents the difference between the vertical coordinates of the (i+2)th and (i+1)th points in the corrected coordinate set, indicating the change in vertical distance between subsequent adjacent points; i+1 -y′ i (x′) represents the difference between the (i+1)th and ith points in the corrected coordinate set, indicating the change in vertical distance between the current two adjacent points; u+2 -x′ u+1 ) is the difference between the horizontal coordinates of the (i+2)th point and the horizontal coordinates of the (i+1)th point in the corrected coordinate set.

[0219] For straight segments, the lengths of all straight segments are summed to obtain the total length and width of the board; for arc segments, the average value of the fillet radius is taken as the overall fillet feature, specifically including:

[0220] For each line segment, calculate its length, assuming the line segment originates from the point (x′). i ,y′ i ) to (x′ j ,y′ j If it is composed of ), then the length L is:

[0221]

[0222] Add up the lengths of all straight line segments to obtain the total length of the board (assuming the board is rectangular, the total length is the sum of the lengths of all horizontal straight line segments, and the total width is the sum of the lengths of all vertical straight line segments); for each arc segment, calculate its corner radius, assuming the arc segment is formed by three points (x′). i ,y′ i ), (x′ i+1 ,y′ i+1 ), (x′ i+2 ,y′ i+2 If the radius of the fillet is determined, the radius R can be calculated by fitting a circle, taking the average value of the fillet radii of all arc segments as the overall fillet feature.

[0223] Calculate the length and width deviations based on the total length and width of the board; calculate the corner radius deviation based on the overall corner characteristics, specifically including: assuming the theoretical length of the board is L. theory The theoretical width is W theory Then the length deviation ΔL and the width deviation ΔW are respectively:

[0224] ΔL=|L actual -L theory |;

[0225] ΔW=|W actual -W theory |;

[0226] Among them, L actual and W actual These are the actual total length and total width of the board, respectively. Assume the theoretical corner radius of the board is R. theory Then the fillet radius deviation ΔR is:

[0227] ΔR=|R average -R theory |;

[0228] Among them, R average It is the average value of the fillet radii of all the arc segments.

[0229] Calculate the corresponding relative deviations of length, width, and fillet radius based on the length deviation, width deviation, and fillet radius deviation, respectively. Specifically, the relative deviation of length, Re, is:

[0230] The relative width deviation Rl is: The relative deviation Ra of the fillet radius is:

[0231] The relative deviations in length, width, and fillet radius are normalized, and the weighted sum of the normalized deviations is calculated to generate a dimensional deviation score.

[0232] In this embodiment of the invention, by calculating the curvature between adjacent points and identifying curvature abrupt change points, the coordinate set can be accurately divided into straight line segments and arc segments. This segmentation method helps to more accurately calculate the straight line length and arc radius of the board, thereby improving the accuracy of dimensional measurement. Calculating the deviation based on the total length, width, and fillet radius of the board can intuitively reflect the difference between the board size and the theoretical value. Summing up the lengths of all straight line segments to obtain the total length and width of the board can comprehensively reflect the linear dimensional characteristics of the board. Taking the average value of the fillet radius of the arc segments as the overall fillet feature can simplify the processing of fillet dimensions while retaining the main information of the fillet dimensions. Calculating relative deviations in length, width, and fillet radius normalizes these deviations to a relative scale, facilitating comparison and evaluation of deviations between boards of different sizes. A weighted summation of these normalized deviations generates a dimensional deviation score, comprehensively reflecting the overall dimensional deviation of the boards and providing a quantitative indicator for quality grading and screening. This score allows for the timely detection of problems in the production process, such as equipment wear and improper process parameters, enabling timely adjustments and improved efficiency. The dimensional deviation score serves as a crucial quality control indicator, monitoring the production quality of boards and ensuring product compliance with design requirements. Furthermore, comparing dimensional deviation scores across different batches or production lines assesses the stability and consistency of the production line. This process can be integrated into automated inspection systems to automatically measure and assess board dimensions, improving inspection efficiency and accuracy. Based on the dimensional deviation score, intelligent decision-making systems can be developed to automatically determine the quality grade of boards and provide corresponding processing suggestions, such as rework, scrapping, or downgrading.

[0233] In a preferred embodiment of the present invention, based on the dimensional deviation score, boards that do not meet the requirements are marked, and the equipment is controlled to perform corresponding processing, including:

[0234] Set a scoring threshold. If the board score exceeds the threshold, it is deemed unqualified. A mark, deviation type, and score value are added to the board surface. The marked board is then transferred to the rework station, where preset cutting, grinding, or shaping parameters are invoked, specifically including:

[0235] Based on the quality requirements and industry standards of the board material, a threshold for dimensional deviation scoring is set, for example, a threshold of 0.1 (assuming the scoring range is between 0 and 1, with larger values ​​indicating larger deviations). This threshold is input into the automated inspection and scoring system as the basis for determining whether the board material is qualified. Dimensional deviation scoring is performed on each board material. The scoring process includes calculating length deviation, width deviation, and corner radius deviation, and then normalizing and weighting the results (as described above). The calculated dimensional deviation score is compared with the set threshold. If the score > the threshold, the board material is deemed unqualified. For unqualified boards, a label containing the deviation type (e.g., length deviation, width deviation, corner radius deviation) and the score value is generated. The process involves using inkjet printers, laser marking machines, or other marking devices to add marking information to the surface of the board. Sensors or vision recognition systems are then used to identify the marked boards, which are then transported to a designated rework station via conveyor belts, robotic arms, or other automated equipment. At the rework station, scanning equipment or a vision system reads the deviation type markings on the board surface. Based on the deviation type, preset cutting, grinding, or shaping parameters are invoked. For example, if the deviation type is length deviation, preset cutting parameters are used to adjust the length; if the deviation type is corner radius deviation, preset shaping parameters are used to correct the corner radius. The board is then reworked using cutting, grinding, or shaping equipment according to the invoked parameters.

[0236] In specific applications, it includes:

[0237] After dimensional deviation scoring, a certain board piece scored 0.18 (greater than the set scoring threshold of 0.1), and was therefore deemed unqualified. The deviation types were length deviation and corner radius deviation. Specifically, the actual length of the board piece was 8mm longer than the theoretical length, and the corner radius was 2mm smaller than the theoretical radius. The parameter adjustment process is as follows:

[0238] A mark is added to the surface of the board: "Length deviation 8mm, corner radius deviation -2mm, score 0.18"; the board is conveyed by a conveyor belt to a comprehensive rework station that includes cutting and straightening equipment. The rework station scans the mark on the surface of the board and identifies the length deviation as 8mm.

[0239] Call the preset cutting parameters: the cutting length is 8mm, the cutting position is one end of the board (determined according to the deviation direction), use the cutting equipment to cut the board according to the called parameters, remove the excess 8mm length, and the rework station continues to identify the corner radius deviation as -2mm.

[0240] Call the preset calibration parameters: increase the calibration radius by 2mm, and the calibration position is all the rounded corners of the board.

[0241] Perform the calibration operation:

[0242] Using a calibration device, the rounded corners of the board are calibrated according to the parameters called in, so that the rounded corner radius reaches the theoretical value. After rework, the board is inspected again to confirm that the length deviation and rounded corner radius deviation have been eliminated. After passing the inspection, the board is released to enter the next process.

[0243] In this embodiment of the invention, by setting a scoring threshold, boards with dimensional deviations exceeding the allowable range can be accurately identified, preventing defective products from flowing into the next process or the final product. Markings, deviation types, and scoring values ​​are added to the board surface, allowing subsequent processing personnel to quickly understand the board's problems and take targeted rework measures. The marked boards are automatically conveyed to the rework station, and preset cutting, grinding, or shaping parameters are invoked to automate the rework process, improving processing efficiency. Automated processing reduces the workload of manual identification and rework, lowers the risk of human error, and improves the overall operating efficiency of the production line. Timely rework of defective boards reduces scrap rates and waste of raw materials and energy. Automated rework allows equipment to operate more efficiently, reducing equipment downtime caused by waiting for manual processing. The markings and scoring values ​​on the board surface provide detailed quality data, facilitating subsequent quality traceability and analysis, and helping to identify potential problems in the production process. Through the analysis and processing of defective boards, the production process and quality control system can be continuously optimized, improving the overall quality level of the product. Strict dimensional deviation scoring and rework ensure the stability and consistency of product quality, improving customer satisfaction.

[0244] Embodiments of the present invention also provide an intelligent control system for a fully automatic edge banding equipment for sheet metal, comprising:

[0245] The determination module is used to construct an octree structure and map pixel values ​​based on the original board image resolution and preset octree depth. The nodes are decomposed into subband sets through multiple filters. The high-frequency subbands are morphologically processed to obtain a binarized edge map. After determining the edge points, Hough transform is used to connect them into curves and interpolation is used to extract the super-resolution edge coordinates.

[0246] The calculation module is used to perform median filtering on the super-resolution edge coordinates and sort them clockwise to form an ordered set. It calculates the curvature markers to divide the edge segments, randomly selects points for each segment as a unit to fit a circle and optimizes it to obtain the final arc model. The sampling points along the arc model are mapped back to the original image coordinate system and merged to form the fitted super-resolution edge coordinate set.

[0247] The normalization module is used to acquire the real-time status parameters of the edge banding equipment, and to normalize the status parameters to obtain a correction factor for correcting measurement deviations.

[0248] The correction module is used to correct the fitted super-resolution edge coordinate set by a correction factor to obtain the corrected edge coordinate set.

[0249] The evaluation module is used to calculate the actual geometric dimensions of the board and the deviation of the actual geometric dimensions based on the corrected edge coordinate set, and to obtain the dimensional deviation score.

[0250] The control module is used to mark boards that do not meet the requirements based on dimensional deviation scores and control the equipment to perform corresponding processing.

[0251] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0252] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0253] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles 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 fully automatic edge banding equipment for sheet metal, characterized in that, The method includes: Based on the original plate image resolution and the preset octree depth, an octree structure is constructed and pixel values ​​are mapped. The nodes are decomposed into subband sets through multiple filters. The high-frequency subbands are morphologically processed to obtain a binarized edge map. After determining the edge points, Hough transform is used to connect them into curves and interpolation is used to extract the super-resolution edge coordinates. After median filtering of the super-resolution edge coordinates, they are sorted clockwise to form an ordered set. Curvature markers are calculated to divide the edge segments. Points are randomly selected for each segment to fit a circle and optimize it to obtain the final arc model. Sampling points along the arc model are mapped back to the original image coordinate system and merged to form the fitted super-resolution edge coordinate set. The real-time status parameters of the edge banding equipment are obtained, and the status parameters are normalized to obtain the correction factor used to correct the measurement deviation. The fitted super-resolution edge coordinate set is corrected by a correction factor to obtain the corrected edge coordinate set. Based on the corrected edge coordinate set, calculate the actual geometric dimensions of the board and the deviation of the actual geometric dimensions, and obtain the dimensional deviation score; Based on the dimensional deviation score, boards that do not meet the requirements are marked, and the equipment is controlled to perform corresponding processing.

2. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 1, characterized in that, An octree structure is constructed based on the original board image resolution and a preset octree depth, and pixel values ​​are mapped. The nodes are then decomposed through multiple layers of filters to obtain a set of subbands, including: Based on the resolution of the original board image and the preset octree depth, an octree structure is constructed, where each node in the octree structure represents a pixel block in the original board image. The pixel values ​​of the original board image are mapped block by block to the nodes of the octree structure; For each node of the octree, convolution operations are performed using a low-pass filter and a high-pass filter of Haar wavelet, respectively, generating one low-frequency subband and three high-frequency subbands after each decomposition. The low-frequency subbands obtained from the current layer decomposition are used as input for the next layer decomposition. The data of the low-frequency subbands are redistributed to the nodes of the octree, and the node size is adjusted to match the resolution after the current decomposition. The filtering operation is repeated on the remapped low-frequency subbands to generate new low-frequency and high-frequency subbands. If the current layer number has not reached the preset layer number, the layer number is incremented and the filtering operation is repeated. If the preset layer number has been reached, the decomposition is terminated to obtain a set of subbands after multi-layer decomposition. The subband set includes the low-frequency subbands of the final layer and the high-frequency subbands of each layer.

3. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 2, characterized in that, Morphological processing of the high-frequency subband yields a binarized edge map. After determining the edge points, Hough transform is used to connect them into a curve, and interpolation is used to extract the super-resolution edge coordinates, including: Morphological processing is performed on the high-frequency subband, and edge points are determined by threshold segmentation in the morphologically processed high-frequency subband. The edge points are input into the Hough transform to identify collinear points and fit them into a continuous curve. Adjacent curve segments are merged to form a continuous edge curve. On a continuous edge curve, linear interpolation is used to obtain the interpolated coordinates of the edge points; Extract the coordinates of the interpolated edge points to form a super-resolution edge coordinate set.

4. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 3, characterized in that, After median filtering of the super-resolution edge coordinates, they are sorted clockwise to form an ordered set. Curvature markers are then used to divide the edge segments, including: Median filtering is applied to the super-resolution edge coordinates, and the filtered coordinate points are sorted in a clockwise direction according to the edge curve to form an ordered edge coordinate set. The local curvature of ordered coordinate points is calculated using cubic spline interpolation. All points are traversed, and points whose curvature exceeds a threshold are marked as segmentation points. The edge is divided into multiple continuous segments based on the dividing point.

5. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 4, characterized in that, By randomly selecting points on each segment as a unit to fit a circle and optimizing it, the final arc model is obtained. The sampled points along the arc model are mapped back to the original image coordinate system and merged to form the fitted super-resolution edge coordinate set, including: Extract the set of coordinate points within each segment as an independent fitting unit; Based on the set of coordinate points of each segment, three points are randomly selected from the coordinate points of the current segment, and the center and radius of the fitted circle are calculated. For all points within the segment, the distance from the current arc model is calculated. Points with a distance less than the threshold are marked as interior points. The center and radius of the circle are re-optimized using the least squares method with all interior points to obtain the final arc model of each segment. For the final circular arc model of each segment, sampling points are uniformly sampled along the circular arc path with a super-resolution step size; The sampling points are mapped back to the original image coordinate system, and the dense coordinate points of all segments are combined in the original sorting order to form the fitted super-resolution edge coordinate set.

6. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 5, characterized in that, The real-time status parameters of the edge banding equipment are acquired, and the status parameters are normalized to obtain correction factors for correcting measurement deviations, including: Obtain the current status parameters of the edge banding equipment, including roller wear, temperature, pressure, motor speed, and sensor calibration error. Each state parameter is normalized to obtain the normalized state parameters; Based on the normalized state parameters, determine the center value and sensitivity of the Gaussian function corresponding to each state parameter; based on the center value and sensitivity of the Gaussian function corresponding to each state parameter, calculate the Gaussian weight of each state parameter, and normalize the Gaussian weight of each state parameter to obtain the normalized Gaussian weight of each state parameter. For each state parameter, calculate the corresponding mean and standard deviation from historical data; Calculate the coefficient of variation based on the mean and standard deviation of each state parameter, and use the coefficient of variation to calculate the initial weight of each state parameter. The normalized Gaussian weight values ​​are multiplied by the corresponding initial weights to obtain the final weights for each state parameter. The correction factor is obtained by multiplying each normalized state parameter by its corresponding final weight and summing the results.

7. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 6, characterized in that, The fitted super-resolution edge coordinate set is corrected by a correction factor to obtain the corrected edge coordinate set, including: Iterate through each point in the fitted coordinate set, and multiply the horizontal coordinate (x) and vertical coordinate (y) of each point by a correction factor to calculate the corrected coordinates.

8. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 7, characterized in that, Based on the corrected edge coordinate set, the actual geometric dimensions of the board and the deviation of the actual geometric dimensions are calculated, and a dimensional deviation score is obtained, including: Calculate the curvature between adjacent points on the corrected coordinate set, identify curvature abrupt change points, and divide the coordinate set into straight line segments and circular arc segments according to the curvature threshold; For straight segments, the lengths of all straight segments are summed to obtain the total length and width of the board; for arc segments, the average value of the fillet radius is taken as the overall fillet feature. Calculate the length and width deviations based on the total length and width of the board; calculate the corner radius deviations based on the overall corner radius characteristics. Calculate the corresponding relative deviations of length, width, and fillet radius based on the length deviation, width deviation, and fillet radius deviation, respectively. The relative deviations in length, width, and fillet radius are normalized, and the weighted sum of the normalized deviations is calculated to generate a dimensional deviation score.

9. The intelligent control method for a fully automatic edge banding equipment for sheet metal according to claim 8, characterized in that, Based on the dimensional deviation score, boards that do not meet the requirements are marked, and the equipment is controlled to perform corresponding processing, including: Set a scoring threshold. If the board score is greater than the scoring threshold, it is judged as unqualified, and a mark, deviation type and score value are added to the board surface. The marked material is transferred to the rework station, where preset cutting, grinding, or shaping parameters are called.

10. An intelligent control system for a fully automatic edge banding equipment for sheet metal, characterized in that, The system performs the method as described in any one of claims 1 to 9, comprising: The determination module is used to construct an octree structure and map pixel values ​​based on the original board image resolution and preset octree depth. The nodes are decomposed into subband sets through multiple filters. The high-frequency subbands are morphologically processed to obtain a binarized edge map. After determining the edge points, Hough transform is used to connect them into curves and interpolation is used to extract the super-resolution edge coordinates. The calculation module is used to perform median filtering on the super-resolution edge coordinates and sort them clockwise to form an ordered set. It calculates the curvature markers to divide the edge segments, randomly selects points for each segment as a unit to fit a circle and optimizes it to obtain the final arc model. The sampling points along the arc model are mapped back to the original image coordinate system and merged to form the fitted super-resolution edge coordinate set. The normalization module is used to acquire the real-time status parameters of the edge banding equipment, and to normalize the status parameters to obtain a correction factor for correcting measurement deviations. The correction module is used to correct the fitted super-resolution edge coordinate set by a correction factor to obtain the corrected edge coordinate set. The evaluation module is used to calculate the actual geometric dimensions of the board and the deviation of the actual geometric dimensions based on the corrected edge coordinate set, and to obtain the dimensional deviation score. The control module is used to mark boards that do not meet the requirements based on dimensional deviation scores and control the equipment to perform corresponding processing.

Images