A cutting and printing device and method with multiple intelligent positioning

Abstract

This invention discloses a multi-stage intelligent positioning cutting and printing method. The method includes steps such as equipment self-inspection, processing mode selection, material adsorption and height adjustment, image acquisition, contour extraction and cutting path generation, target contour cutting, printing path alignment correction, pattern printing, defect detection, and finished product rejection. It employs multi-round image scanning combined with markerless recognition technology to achieve highly robust alignment control of material posture and boundaries. The device integrates a control system, adsorption platform, integrated camera, lifting device, cutting device, UV printer head, and cleaning device, supporting full-process image acquisition and execution path coordination. Consistent matching between the cutting path and pattern path is achieved by fusing Hu moment, curvature features, and ICP registration, and dynamic printing compensation is performed using a PID control strategy, constructing a closed-loop control mechanism for cutting and printing. This invention improves image and text alignment accuracy and processing automation level, and is suitable for flexible customized production scenarios.

Description

Technical Field

[0001] This invention relates to the field of precision material processing technology, and in particular to a cutting and printing device and method with multiple intelligent positioning. It is especially suitable for integrated UV curing printing and CNC cutting of sheet and roll materials in industries such as advertising decoration, packaging processing, and hardware products. The core technologies encompass cross-disciplinary areas such as machine vision positioning, CNC cutting, UV printing, and automated collaborative control. Background Technology

[0002] With the development of intelligent manufacturing and flexible processing technologies, the demand for personalized pattern output and precise graphic cutting is increasing in industries such as advertising production, clothing customization, and signage processing. Traditional processing methods typically separate pattern printing and outline cutting into two independent stages, completed separately by printing and cutting equipment. This separate process not only increases intermediate steps such as manual handling and material positioning, easily leading to problems like graphic misalignment and outline deviation, but also results in low processing efficiency and unstable alignment accuracy, making it difficult to meet the demands of high-efficiency, high-precision, and flexible mass production.

[0003] Some existing systems attempt to improve the automation level of processing by incorporating image recognition functions, but these are mostly limited to single functional links (such as cutting path auxiliary generation, printing layer correction, etc.) and fail to achieve complete closed-loop control from image acquisition, contour extraction, cutting path planning to pattern output. Although some systems have basic camera modules and image recognition capabilities, they generally rely on manually set physical markers (such as positioning crosshairs, border barcodes, etc.) for alignment recognition, lacking the ability to adaptively recognize and correct for actual material deformation and contour changes in real time.

[0004] Furthermore, existing equipment generally does not integrate scanning, cutting, and printing on the same operating platform, lacking unified scheduling and path coordination capabilities. This often results in significant deviations between the cutting path and the printed pattern, especially when dealing with high-frequency batch changes and small-batch customized processing, making it difficult to guarantee system stability and finished product yield. Although some systems propose solutions that combine scanned images with execution control paths, these solutions mostly remain at the level of offline calibration or single path correction, and have not yet achieved multi-round visual feedback and dynamic compensation control throughout the entire processing process.

[0005] Therefore, there is an urgent need for a new cutting and printing method and device that integrates image perception, multi-axis execution control and path dynamic correction capabilities to achieve closed-loop control of the entire process from image recognition, graphic cutting to pattern output, and effectively improve the consistency of images and text, process stability and equipment integration level. Summary of the Invention

[0006] This invention aims to solve problems such as the separation of pattern printing and contour cutting, process breakage, reliance on manual marking for positioning, frequent misalignment of images and text, and lack of a unified control platform in existing processing equipment. It proposes an intelligent processing method and device that integrates cutting, scanning and printing, and has multi-round image recognition and dynamic path compensation capabilities, so as to achieve high-precision and automated cutting-printing consistency control of materials under unmarked conditions.

[0007] To address the aforementioned technical problems, the present invention provides, in one aspect, a multi-stage intelligent positioning cutting and printing method, wherein the method includes:

[0008] Step 1: Start the equipment and complete the self-test of the cutting device, lifting device, integrated camera, cleaning device, UV printer head and adsorption platform.

[0009] Step 2: Select the cutting mode, printing mode, or cutting and printing mode according to the processing task, and initialize the corresponding control parameters;

[0010] Step 3: Start the negative pressure adsorption of the adsorption platform to fix the material on the adsorption platform, and adjust the imaging height of the integrated camera by the lifting device according to the ultrasonic thickness measurement results;

[0011] Step 4: Control the crossbeam to drive the integrated camera to perform the first scan, acquire the initial image of the material, and transmit it to the computer system;

[0012] Step 5: Extract material contours based on the initial image, combine them with a preset cutting template to complete feature matching, generate a cutting path, and drive the cutting device to complete the cutting;

[0013] Step 6: After cutting, perform cleaning and a second scan to obtain the actual contour image, and use this image to dynamically match and align the printing template.

[0014] Step 7: Drive the UV printer head to print the pattern along the corrected path. During this process, a third scan is performed and an image is acquired to detect material offset. If the deviation exceeds the threshold, the printing path parameters are adjusted to achieve dynamic compensation for printing alignment.

[0015] Step 8: After printing is completed, perform a fourth scan and acquire images. Based on grayscale deviation and contour position deviation, identify and mark missing prints, misalignments, or cutting defects.

[0016] Step 9: Remove defective products according to the defect markings and clean up the platform residue to proceed to the next processing cycle.

[0017] Optionally, in step 5, by formula 27

[0018]

[0019] The system performs similarity calculation between the printed template outline and the actual outline after cutting. It fuses the Hu moment feature distance and curvature feature distance according to a weighted coefficient to obtain a unified similarity index Sim. When Sim is lower than the preset matching threshold, the system triggers local deformation correction based on thin plate spline interpolation (TPS) and global path alignment based on ICP registration, thereby realizing adaptive adjustment of the printing path to the cutting outline.

[0020] Optionally, in step 7;

[0021] Through formulas 36 to 38

[0022]

[0023] Real-time detection of X-axis deviation ΔX, Y-axis deviation ΔY, and overall deviation Δ during the printing process;

[0024] Using formulas 39 and 40

[0025]

[0026] The system performs dynamic path compensation for the aforementioned deviations. The compensation amount is calculated based on a PID control algorithm and includes three control parameters: proportional response, integral adjustment, and derivative correction. These parameters are used to correct the motion path of the UV printer head in real time, ensuring that the printed pattern is aligned with the cutting contour with high precision under unmarked conditions.

[0027] Optionally, in step 6, by formula 35

[0028]

[0029] The SSD squared error calculation operation shown determines the degree of spatial deviation between the printed template outline and the actual outline after cutting. When the SSD error value exceeds the set tolerance threshold, the control system triggers the recalibration logic of the printing path to improve the alignment accuracy and avoid pattern misalignment.

[0030] Optionally, in step 5, formulas 16 to 26 are used respectively.

[0031] ;

[0032] ;

[0033] ;

[0034] ;

[0035] ;

[0036] ;

[0037] ;

[0038] ;

[0039] ;

[0040] ;

[0041]

[0042] The Hu moment feature extraction and curvature feature extraction operations of the material contour are implemented to jointly describe the global shape and local variation features of the target graphic. The feature values ​​serve as the basic input for subsequent template matching and path planning, effectively improving the matching robustness and alignment stability in the unmarked environment.

[0043] Optionally, when the control system is set to cutting mode, the method only executes steps 1 to 5, without starting the cleaning device, image correction, or printing. After completing the multi-axis linkage cutting operation of image acquisition, template matching, and target contour, the current processing cycle ends.

[0044] Optionally, when the control system is set to printing mode, the method only executes the process of steps 1, 2, 3 and 7, and directly starts the pattern printing operation based on the positioning relationship between the material outline and the printing template, omitting the outline cutting, cleaning and secondary scanning steps, which is suitable for finished product processing tasks that do not require structural cutting.

[0045] To address the aforementioned technical problems, another aspect of the present invention provides a cutting and printing device with multiple intelligent positioning, comprising:

[0046] The control system is used to uniformly schedule the various execution modules and run image processing and path planning algorithms.

[0047] The adsorption platform has a controllable negative pressure adsorption function and is used to fix the material to be processed.

[0048] An integrated camera is used to acquire original images and cut images before and after processing to achieve multi-stage image recognition and position correction;

[0049] A lifting device is used to adjust the imaging height of the integrated camera according to the thickness of the material;

[0050] The cutting device is installed on a three-axis linkage platform and performs material cutting according to the path generated by the control system.

[0051] The UV printer head is configured to print patterns on materials along a preset path and supports image feedback control.

[0052] A cleaning device is used to remove debris after cutting, ensuring image quality for both image acquisition and printing; among which...

[0053] The control system is configured to support the following functional processes:

[0054] After the device is initialized, the first image acquisition and contour extraction of the material are completed. The matching algorithm is executed to generate the cutting path and drive the cutting device to complete the cutting operation. Then, the image is acquired again to correct the printing template and the deviation detection and path dynamic compensation control are executed to finally achieve the consistency of the cutting-printing closed loop.

[0055] Furthermore, it supports coordinated scheduling and closed-loop data control of each stage in the processing flow, including image acquisition, contour extraction, path generation, image correction, and deviation compensation, to ensure that the device achieves high-precision consistency between the cutting path and the printed pattern under unmarked conditions.

[0056] Optionally, the apparatus is adapted to perform only image-guided cutting processes and includes:

[0057] An integrated camera for capturing raw images of materials;

[0058] A lifting device is used to adjust the imaging height of the camera;

[0059] The cutting device, installed on a three-axis linkage platform, is used to perform path cutting after image recognition;

[0060] The control module is used to extract image contours, perform template matching and path planning, and control the cutting device to perform precise cropping; wherein...

[0061] The control module supports image alignment and path generation based on contour recognition and feature matching under unmarked conditions, realizing image-execution closed-loop control in pure cutting mode.

[0062] Optionally, the apparatus is adapted to perform only a pattern printing process and includes:

[0063] An integrated camera for capturing images of positioned materials;

[0064] A UV printer head used to print patterns along a set path;

[0065] The control module is used to compare the image with the print template and correct the print path; among which...

[0066] The control module supports dynamic correction of the printing path based on SSD error and position offset calculation results, and performs continuous image acquisition and feedback compensation during the printing process to achieve accurate pattern output in pure printing mode.

[0067] The beneficial effects of the technical solution of this invention are:

[0068] The multi-position intelligent cutting and printing device of this invention enables high-precision alignment control throughout the entire process. Through four image acquisitions and recognitions—before cutting, after cutting, after the first print, and after printing—multiple intelligent visual positioning is achieved, thus constructing a closed-loop processing flow of "cutting—scanning—printing." This technology overcomes the reliance on manual marking and fixed fixtures in traditional cutting and printing processes, significantly improving alignment accuracy and flexible processing capabilities, making it particularly suitable for scenarios with irregular materials or large batch deviations.

[0069] The multi-positioning intelligent cutting and printing device of the present invention can improve the robustness of matching the printed pattern with the cutting contour. By integrating the contour matching method of Hu moment and curvature features, and introducing the unified similarity index Sim defined by Formula 27, the device combines the thin plate spline (TPS) and iterative nearest point (ICP) algorithm to achieve global and local path adaptive correction of cutting error, effectively avoiding the risk of misalignment between the pattern and the actual contour, and significantly improving the robustness and adaptability of printing alignment.

[0070] The cutting and printing device with multiple intelligent positioning of the present invention can support dynamic path compensation, improve real-time performance and production line adaptability. During the printing process, the UV print head path is dynamically compensated by real-time image acquisition and deviation detection algorithm (formulas 36-38) based on PID control model (formulas 39-40), so that the pattern can still be accurately output even when the material is offset or deformed, meeting the real-time response requirements of high-speed production line operation.

[0071] The multi-positioning intelligent cutting and printing device of the present invention can construct a judgment and rejection mechanism to improve the yield of finished products. By introducing a fourth image acquisition after printing and calculating grayscale deviation and contour position deviation, the system can accurately identify defects such as missing print, misalignment or cutting deviation, and execute the rejection logic of non-conforming products based on the judgment result, thereby realizing closed-loop control of finished product quality and significantly improving the overall yield.

[0072] The multi-position intelligent cutting and printing device of the present invention can support decoupled operation of cutting mode and printing mode, improve system adaptability and flexibility, and provide three switchable operation modes: cutting mode, printing mode and cutting and printing mode. The execution process and control logic can be flexibly selected according to actual production needs, supporting pure cutting or pure printing process, realizing modular decoupling and multi-task compatibility of the system, and broadening the applicability of the equipment in different processing scenarios. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the steps of the cutting and printing method with multiple intelligent positioning in an embodiment of the present invention;

[0074] Figure 2 This is a perspective view of the cutting and printing device with multiple intelligent positioning in an embodiment of the present invention;

[0075] Figure 3 This is a schematic diagram of the installation of the integrated camera in an embodiment of the present invention;

[0076] Figure 4 This is a schematic diagram of the lifting device in an embodiment of the present invention;

[0077] Figure 5 This is a schematic diagram of the cutting device in an embodiment of the present invention;

[0078] Figure 6 for Figure 3 Enlarged view of point A in the middle;

[0079] Figure 7 This is a schematic diagram of the integrated camera structure in an embodiment of the present invention;

[0080] Figure 8 This is an enlarged schematic diagram of the integrated camera lens in an embodiment of the present invention. Detailed Implementation

[0081] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0082] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0083] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0084] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0085] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0086] Please see Figure 2 , Figure 3 , Figure 4 and Figure 6 As shown, a system architecture of one embodiment is illustrated:

[0087] A multi-position intelligent cutting and printing system includes a UV printer frame 1, with a horizontally extending crossbeam 8 slidably connected to the UV printer frame 1 in the front-to-back direction. A UV printer head 2 is mounted on the crossbeam 8 and is slidably connected to the crossbeam 8 in the left-to-right direction. The UV printer head 2 can move vertically. An adsorption platform 7 is mounted on the UV printer frame 1, and the adsorption platform 7 has a negative pressure adsorption function. A lifting device 4 is mounted on the crossbeam 8 above the adsorption platform 7. An integrated camera 5 is mounted on the lifting device, and a cleaning device 6 is mounted on the lifting device 4 in front of the integrated camera. A cutting device 3 is mounted on the UV printer head 2.

[0088] It also includes a computer and several sensors connected to the computer. The UV printer head 2, cutting device 3, lifting device 4, integrated camera 5, and cleaning device 6 are electrically connected to the computer. The sensors include an ultrasonic sensor installed on the UV printer frame 1 for detecting the thickness of the material, an external trigger sensor installed on the UV printer frame 1 and located at the rear end of the adsorption platform, and multiple limit sensors for limiting the position.

[0089] The lifting device 4 includes lifting modules fixed on both sides of the camera. The two lifting modules are fixedly connected by a reinforcing profile 411. The lifting module includes a vertically extending double guide rail lead screw module 401. A stepper motor 403 is installed on the double guide rail lead screw module 401. The stepper motor is driven by a reducer 402. The reducer 402 is driven by the lead screw of the double guide rail lead screw module 401. A slider connecting plate 404 is installed on the slide of the double guide rail lead screw module 401. There are two bolt slots extending in the left and right direction on the slider connecting plate 404. The two bolt slots are arranged vertically in parallel. The slide and the slider connecting plate 404 are connected by bolts. The installation position of the slider connecting plate 404 can be changed according to the position of the bolt in the bolt slot.

[0090] A profile fixing plate 405 is fixedly connected to the side of the slider connecting plate 404 near the integrated camera 5. A fixing profile 406 extending in the front-back direction is installed on the profile fixing plate 405. A profile nut is installed inside the fixing profile 406. By changing the front-back position of the profile nut inside the fixing profile 406, the position of the fixing profile 406 fixed on the profile fixing plate 405 is changed.

[0091] A camera mounting plate 409 is fixedly connected to the rear end of the fixed profile 406. The inner side of the camera mounting plate 409 is fixedly connected to the integrated camera 5. The left and right ends of the reinforcing profile 411 are fixedly connected to the inner sides of the two fixed profiles 406 respectively. A machine mounting plate 408 is fixedly connected to the inner side of the double guide rail screw module 401. The machine mounting plate 408 extends in the front-back direction. Multiple sets of bolt holes are opened on the machine mounting plate 408 in the front-back direction. The front-back position of the machine mounting plate 408 can be changed by using bolt holes in different positions. A module nut is installed on the inner side of the double guide rail screw module 401. The vertical position of the machine mounting plate 408 can be changed by changing the vertical position of the module nut.

[0092] A reinforcing angle aluminum 407 for strengthening structural stability is also provided between the machine fixing plate 408 and the double guide rail screw module 401. A machine adapter plate 410 is fixedly connected to the outer rear end of the machine fixing plate 408. The crossbeam is slidably connected to the UV printer frame 1 through the vertically extending crossbeam bracket. The upper end of the crossbeam bracket is fixedly connected to the crossbeam 8, and the lower end is slidably connected to the UV printer frame 1 in the front-back direction. The machine adapter plate 410 is fixedly connected to the crossbeam bracket.

[0093] like Figure 5As shown, the cutting device includes a blade holder fixing plate 34 fixed to the UV printer head 2. A blade holder 33 is slidably mounted on the blade holder fixing plate in the vertical direction. A blade holder screw is rotatably connected to the blade holder. A blade holder motor 36 is fixed to the blade holder fixing plate 34. The blade holder screw is driven by the blade holder motor 36, thereby driving the blade holder 33 to slide up and down on the blade holder fixing plate 34. A blade head motor 35 is mounted on the blade holder. The blade head motor 35 is driven by a drive gear. A blade head 31 coaxial with the drive gear is installed inside the drive gear. When the blade head motor 35 rotates, it drives the drive gear to rotate, thereby driving the blade head 31 to rotate. A punching device 32 is also mounted on the blade holder. The punching device includes a punching cylinder and a waste collection box fixed to the blade holder, a punching rod fixed to the movable end of the punching cylinder, and a punch fixed to the lower end of the punching rod. The punching rod and the punch are hollow. The upper end of the punching rod is connected to the waste collection box, and the lower end of the punching rod is connected to the punch.

[0094] like Figure 7 and Figure 8 As shown, the cleaning device includes an inner brush clamp 64 located in front of the integrated camera 5. An outer brush clamp 65 is fixed to the inner brush clamp 64 by bolts. A brush 66 extending in the left-right direction is installed between the inner brush clamp 64 and the outer brush clamp 65. Tightening the bolts can fix the brush between the inner brush clamp 64 and the outer brush clamp 65. A brush fixing plate 63 is fixed to each of the left and right ends of the inner brush clamp 64. A cylinder fixing plate 62 is fixed to the camera fixing plate 409. A brush cylinder 61 extending in the up-down direction is installed on the cylinder fixing plate. The movable end of the brush cylinder 62 is fixed to the brush fixing plate 63. The extension and retraction of the brush cylinder 61 can drive the brush to move up and down.

[0095] The integrated camera 5 includes a camera bracket 51, a camera lens 52, a built-in light source 53, and a camera chip. The camera lens 52 is an integrated lens and is fixed on the camera bracket 51. The built-in light source 53 is fixed on the camera bracket 51 and is located on the front and rear sides of the camera lens 52. The camera chip is installed inside the camera bracket 51 and is positioned opposite to the camera lens 52.

[0096] The following are the detailed working steps of the multi-position intelligent positioning cutting and printing device in this embodiment, including:

[0097] Step 1: Turn on the main power supply of the equipment and the computer. The computer will automatically start the control program and perform self-checks on each component: check the left and right sliding and vertical movement functions of the UV printer head 2 to ensure that the print head is not blocked; check whether the rotation of the cutter head 31, the up and down sliding of the cutter seat 33, and the cylinder action of the punching device 32 are normal; verify the smooth operation of the double guide rail lead screw module 401 of the lifting device 4, and the normal response of the stepper motor 403 and the reducer 402; test the extension and retraction function of the brush cylinder 61 of the cleaning device 6 to confirm that the brush 66 is firmly fixed and the brush height can be precisely controlled by the cylinder; check the negative pressure system of the adsorption platform to ensure that the adsorption holes on the platform surface are not blocked and the negative pressure value meets the standard.

[0098] Step 2: Select the corresponding working mode on the computer interface according to production needs. There are three types:

[0099] Cutting mode: Selectable only when material needs to be cut;

[0100] Printing mode: Selectable only when UV coating printing is required on the material;

[0101] Cutting and printing mode: Selected when material cutting and subsequent printing processing need to be completed continuously.

[0102] Step 3: After selecting the cutting and printing mode, the computer controls the crossbeam 8 to move to the rear standby position of the adsorption platform 7; the negative pressure adsorption function of the adsorption platform is activated, and the flexible material to be processed is laid flat on the platform surface; the ultrasonic thickness sensor automatically detects the material thickness and transmits the data to the computer in real time; the computer generates a height adjustment command based on the material thickness and the fixed focal length parameters of the integrated camera 5 and sends it to the stepper motor 403; after the stepper motor 403 starts, it drives the lead screw of the double guide rail lead screw module 401 to rotate through the reducer 402, which drives the slider connecting plate 404 to move vertically; the slider connecting plate 404 is linked to fix the profile 406 through the profile fixing plate 405, and finally the camera fixing plate 409 drives the integrated camera 5 to rise and fall synchronously; the stepper motors 403 on the left and right sides keep working synchronously until the distance between the lower end face of the integrated camera 5 and the upper end face of the material reaches the fixed focal length requirement, and the motor stops running.

[0103] Ultrasonic echo method for thickness measurement, calculating the actual thickness of the material:

[0104] Formula 1:

[0105] Where d is the material thickness (mm), v is the propagation speed of the ultrasonic wave in the corresponding material (mm / s), and t is the time difference between ultrasonic wave transmission and reception (s).

[0106] Adjust the camera height to determine the optimal focal length alignment between the camera and the material:

[0107] Formula 2:

[0108] Where H is the final height of the camera (mm, relative to the platform reference plane). This is the initial reference height for the camera (mm, fixed after equipment calibration). Set a fixed focal length compensation amount (mm) for the camera, such as when the overall lens focal length is 10mm. =8mm).

[0109] Stepper motor drive pulse calculation converts height requirements into motor control signals, precisely controlling the motor rotation angle through the number of pulses, achieving camera height adjustment in the ±0.01mm range.

[0110] Formula 3:

[0111] Where N is the number of pulses required by the stepper motor, P is the lead of the dual-rail lead screw module (mm / revolution), S is the step angle of the stepper motor (° / step), and i is the reduction ratio of the reducer.

[0112] Step 4: The crossbeam 8 moves the integrated camera 5 forward. When the crossbeam 8 passes the external trigger sensor, the integrated camera 5 starts scanning. The crossbeam 8 moves the integrated camera 5 forward evenly until the front limit sensor is triggered. The crossbeam 8 stops moving, and the integrated camera 5 completes the scanning.

[0113] Step 5: The integrated camera 5 transmits the initial scan image to the computer in real time. The computer uses image recognition algorithms to extract precise position information such as material edges and positioning marks, and performs pixel-level alignment matching between the preset cutting template and the actual placement position of the material. After matching, the cutting device 3 is started: the UV printer head 2 drives the cutting device 3 to move left and right (X-axis) along the crossbeam. The crossbeam simultaneously drives the UV printer head 2 and the cutting device 3 to move back and forth (Y-axis) along the UV printer frame 1. At the same time, the blade holder 33 completes the up and down (Z-axis) adjustment on the blade holder fixing plate 34, and cooperates with the blade head motor 35 to drive the blade head 31 to rotate 360° at high speed. Through the XYZ three-axis linkage, the cutting device achieves full coverage of the material plane and precise adaptation of the cutting depth, and completes the processing of the preset cutting path.

[0114] Converting color scanned images to grayscale reduces computational load, while preserving image features through weighted coefficients ensures overall lens imaging consistency.

[0115] Formula 4:

[0116] Among them, R, G, and B are the original pixel's three color channel values.

[0117] Gaussian filtering denoising suppresses environmental reflections and dust interference. Image smoothing is achieved through Gaussian weighted averaging, preserving the edges of location markers.

[0118] Formula 5:

[0119] Where G(x,y) is the pixel value at the filtered coordinates (x,y). =1.2 (filter radius), k=3 (3×3 window size), I(x,y) is the original pixel value. Coordinates in the original image pixel values, This is a Gaussian weighting function.

[0120] Adaptive Histogram Equalization: Image segmentation divides the entire image into small sub-blocks, avoiding local overexposure / underexposure caused by traditional histogram equalization; Sub-block histogram statistics count the number of pixels at each gray level in each sub-block; Contrast limiting prevents overexposure in localized bright areas; CDF mapping stretches the gray-level range of sub-blocks to 0-255, enhancing edge contrast; Bilinear interpolation eliminates abrupt changes in brightness (artifacts) at sub-block boundaries, ensuring image smoothness.

[0121] Formula 6:

[0122] in, The output is the pixel value after CLAHE processing. Let Adj = {(i,j),(i,j+1),(i+1,j),(i+1,j+1)} be the pixel coordinates of the input image, and let Adj = {(i,j),(i,j+1),(i+1,j+1)} be the pixel coordinates. The indexes of its four adjacent sub-blocks, , where is the bilinear interpolation weight for the four adjacent sub-blocks. This refers to the minimum cumulative number of pixels in the corrected histogram. Let be the pixel count of gray level t in the (i,j)th sub-block. To limit the contrast threshold, =3.0~5.0 is the limiting factor, L=256 is the gray level. This refers to the histogram count of clipping values ​​that exceed a threshold. The average amount of excess pixels. For excess total pixels, The grayscale value mapping within the sub-block. , refers to the corrected pixel count of gray level s in the (i,j)th sub-block, and m and n refer to the vertical and horizontal pixel counts of a single sub-block, respectively. This refers to the cumulative number of pixels at gray level 0 in the histogram after the correction of the (i,j)th sub-block. k, s, and t are gray level indices. It iterates through all gray levels from 0 to 255 and is used for histogram statistics and cumulative calculation.

[0123] Material contour extraction, Canny edge detection optimization, calculation of pixel grayscale change rate (gradient), and edge location:

[0124] Formula 7:

[0125] Formula 8:

[0126] Formula 9:

[0127] Formula 10:

[0128] in, G represents the gradients in the x and y directions, respectively (calculated through convolution using the Sobel operator, a commonly used edge detection operator that can highlight edges in the horizontal / vertical directions), and G is the gradient magnitude (reflecting edge strength; the larger the value, the clearer the edge). The gradient direction (reflects the edge direction and provides a basis for subsequent edge screening).

[0129] Adaptive dual thresholds filter valid edges and eliminate weak noise interference:

[0130] Formula 11:

[0131] Formula 12:

[0132] in, The maximum gradient value of the entire image. A high threshold is set (pixels with gradient values ​​exceeding this value are directly considered valid edges). Set a low threshold (pixels with gradient values ​​below this value are directly discarded); reduce the threshold when edges are blurry. (e.g., 0.25), to avoid effective edges being discarded, gradient > To determine the edge; gradient < Remove; retain edges that are in between and connected, supplement weak edges (such as slightly blurred edges of materials) by connectivity judgment, and remove isolated noise points to ensure that the extracted edges are continuous and complete, adapting to the accurate capture of material contours in unmarked scenarios.

[0133] Morphological closing operation:

[0134] Formula 13:

[0135] in For expansion operations, For the erosion operation, K is a 5×5 rectangular kernel. A 5×5 rectangular kernel was chosen because it effectively fills in contour gaps and eliminates burrs without excessively eroding the main contour. The process involves first dilating and then eroding to fill in contour gaps while preserving contour dimensions. (Dilution operation) It will expand the bright area and fill small gaps inside the outline; erosion operation It will shrink the bright area and eliminate burrs on the edge of the outline; after the two are combined, the overall size of the outline remains basically unchanged, only local defects are repaired, and a continuous and complete material outline is obtained.

[0136] Area Filtering:

[0137] Formula 14: ; Keep S≥0.5×W×H, where (W / H) is the minimum preset size of the material in pixels.

[0138] Where S is the pixel area of ​​the outline (the total number of pixels contained in the outline), and (W, H) are the number of pixels corresponding to the minimum size of the material preset before production. The outline with an area ≥ 50% of the minimum preset area is retained to ensure that only the main outline of the material is retained.

[0139] Coordinate output, extract contour point set:

[0140] Formula 15:

[0141] The contour chain is formed by clockwise sorting, which transforms the contour into an ordered set of coordinate points, providing a data foundation for subsequent feature description and matching. Clockwise sorting ensures that the order of the contour points is consistent with the order of the template contour points, reducing the amount of computation during matching. The contour chain is a continuous coordinate sequence that can completely describe the edge shape of the material.

[0142] Contour feature description (resistant to rotation / scaling / translation, ensuring robustness to markerless matching)

[0143] Hu moment features (7 invariant moments, global features)

[0144] Geometric moments describe the geometric distribution characteristics of the profile and lay the foundation for subsequent calculations of invariant moments.

[0145] Formula 16:

[0146] in, It is the zeroth moment, equal to the pixel area S of the contour; This is a first-order moment used to calculate the centroid of the profile; It is a second moment, describing the degree of dispersion of the profile.

[0147] Central moment, eliminating translational interference (material placement offset does not affect features):

[0148] Formula 17:

[0149] in, The coordinates of the centroid (geometric center) of the profile. The coordinates of a pixel relative to the centroid are given by the central moment, which ensures that the feature values ​​remain unchanged after the material is translated by moving the origin of the coordinate system to the centroid.

[0150] Normalized center distance, excluding scaling interference (minor changes in material size do not affect features):

[0151] Formula 18:

[0152] in, By dividing by The power of the value normalizes the central moment to a fixed range, ensuring that the characteristic value remains unchanged after the material is scaled.

[0153] The Hu moment formula, through the combination of normalized central moments, forms an eigenvector H=[ that is invariant to rotation, scaling, and translation.] The Hu moments are classic global profile features. Regardless of how the material is positioned (translated), how small the size changes (scaled), or how it is rotated, the values ​​of the seven Hu moments remain basically unchanged, serving as a "global identity card" for the material profile.

[0154] Formula 19:

[0155] Formula 20:

[0156] Formula 21:

[0157] Formula 22:

[0158] Formula 23:

[0159] Formula 24:

[0160] Formula 25:

[0161] Contour curvature features are used to supplement local detail features and improve the matching accuracy of irregular contours.

[0162] Formula 26:

[0163] Sample points are taken from the contour chain at 5-pixel intervals (to avoid excessive computation). , , The numerator is the absolute value of the cross product of the two vectors formed by the three points (reflecting the angle between the vectors, i.e. the curvature of the contour), and the denominator is the cube of the magnitude of the first vector (normalized curvature value). The curvature of the i-th sampling point (the larger the value, the more obvious the contour curvature) can capture local convexities, depressions and other details of the material contour, making up for the lack of global features of Hu moments.

[0164] Template outline matching and registration:

[0165] Template preprocessing: Import the vector image of the preset cutting template, and follow the process of "Canny edge detection → morphological closing operation → contour filtering → feature extraction" from step 5 to obtain the Hu moment features of the template contour. Curvature characteristics This ensures that the extraction standards for template features and material features are consistent, avoiding matching deviations caused by different processing procedures, and providing a unified benchmark for subsequent similarity calculations.

[0166] Similarity calculation: Sim represents the similarity score (0 ≤ Sim ≤ 1). A match is considered valid when Sim ≥ 0.9, indicating that the material outline and the template outline have highly consistent global and local features.

[0167] Formula 27:

[0168] in, (Hu moment weights, global features dominate). (Curvature weighting, local feature supplementation); For material Hu, For template Hu, The Hu moment Euclidean distance (the smaller the value, the more similar the global features); For the curvature of the material, For template curvature, is the Euclidean distance of curvature (the smaller the value, the more similar the local features); (Maximum possible distance of Hu moment, preset constant) =1.0 (maximum possible distance of curvature, preset constant);

[0169] ICP registration optimization (iterative nearest point algorithm, accurately eliminating deviations), iterative nearest point matching: for each point pϵContour in the material contour point set, in the template contour point set... Find Euclidean distance Find the smallest point q, establish corresponding point pairs, and solve for the rotation matrix R and translation vector T using the least squares method to minimize the error. Update coordinates When the error Stop iterating when the sum of squared pixels is reached.

[0170] Three-axis linkage cutting path generation

[0171] Planar coordinate mapping controls the movement trajectory of the cutting device in the X / Y plane (adsorption platform plane):

[0172] Formula 28:

[0173] Formula 29:

[0174] in, Let be the physical coordinates of the cutting device at time t. Starting coordinates This refers to the X / Y axis feed rate.

[0175] S-curve acceleration and deceleration control avoids impact during the start-up and shutdown of the cutting device, protecting the cutter head and materials.

[0176] Formula 30:

[0177] Formula 31:

[0178] Formula 32:

[0179] in, The initial velocity, For maximum acceleration, For acceleration and deceleration time (smooth transition). This represents the maximum feed rate.

[0180] Cutting depth calculation controls the depth of the cutter head, ensuring penetration of the material without damaging the platform.

[0181] Formula 33:

[0182] in, The height of the cutter head. d represents the material coefficient (1.0 for flexible materials, 1.2 for rigid materials, ensuring complete penetration for rigid materials), and d represents the material thickness (unit: mm). This is the cutting allowance.

[0183] When the cutting path includes the "punching mark", the punching device linkage control is activated: → the cutter head rises 5mm (to avoid interference with the punching device) → the punching cylinder is vented (pressure 0.6MPa, to ensure sufficient punching force) → the punch descends d+0.5mm (to completely penetrate the material) → the punch rises and resets → the waste material is sucked into the waste collection box through the hollow punching rod.

[0184] Step 6: After cutting is completed, the crossbeam 8 is located at the rear of the adsorption platform 7. Cutting stops, and the adsorption platform temporarily maintains a negative pressure state. The brush cylinder 61 of the cleaning device 6 is activated, and its movable end extends and drives the brush fixing plate 63 to move down, so that the lower end of the brush 66 is lower than the lens of the integrated camera 5 and fits against the upper surface of the material. The computer controls the crossbeam 8 to move forward, and the brush 66 moves with the crossbeam to sweep the cutting residue remaining on the material surface to the waste area at the edge of the platform. When the crossbeam 8 triggers the external trigger sensor again, the integrated camera 5 starts a second scan. The second scan is completed until the crossbeam 8 triggers the front limit sensor to stop moving. Then the movable end of the brush cylinder 61 retracts, driving the brush 66 to return to above the lower surface of the integrated camera 5.

[0185] The integrated camera 5 transmits the image of the cut material after secondary scanning to the computer. The computer calls the preset printing template and performs a secondary precise matching between the printing template and the material position based on the actual outline of the cut material to ensure that the relative position of the printed pattern and the cutting outline meets the design requirements.

[0186] Contour extraction after cutting: Following the "Canny edge detection → morphological closing operation → contour filtering" process in step 5, the contour point set of the cut material is obtained:

[0187] Formula 34:

[0188] SSD squared difference matching calculates the similarity between the printed template outline and the cut outline:

[0189] Formula 35:

[0190] in, These are the outline points (pixel coordinates) of the printing template. The corresponding point (pixel coordinates) of the cut contour is denoted by SSD, which is the sum of squared differences (the smaller the value, the higher the matching degree). When SSD ≤ 10 (pixel sum of squares), it is determined to be a valid match, that is, the positional deviation between the template and the cut contour is minimal.

[0191] Deformation correction: If there is local deformation (such as a deviation of a single contour point > 0.3 pixels), thin plate spline interpolation (TPS) is used to adjust the contour of the printing template. TPS interpolation can flexibly adjust the local shape of the printing template according to the deformation trend of the contour after cutting, so that the fit between the template and the actual cutting contour is ≥ 99.5%, avoiding misalignment of the printing due to deformation.

[0192] Coordinate mapping: Through the ICP registration logic in step 5, the adjusted printing template path is mapped to the device physical coordinates (X / Y axis) to ensure that the printing path and the cutting path are in the same coordinate system, and the cutting deviation is ≤0.1mm, which meets the product accuracy requirements.

[0193] Step 7: After matching is complete, the UV printer head 2 begins operation. The crossbeam 8 moves the UV printer head backward, slowly completing the printing job. When the crossbeam reaches the rear limit switch, it stops moving backward and begins moving forward. When the crossbeam 8 passes the external trigger sensor, the integrated camera 5 starts scanning again. When the crossbeam reaches the front limit sensor, the crossbeam 8 stops moving, and the integrated camera 5 transmits a signal to the computer. The computer automatically compares the outline of the scanned image with the cut outline of the previous scan. If the deviation exceeds the set threshold, the path parameters for the next print are automatically corrected; if the deviation is within the threshold, the original parameters are used. The UV printer head 2 continues printing according to the corrected (or original) parameters, and this cycle repeats until printing is complete.

[0194] Dynamic printing deviation correction compensates for minute offsets during the printing process in real time, adapting to dynamic changes in unmarked outlines:

[0195] Deviation calculation, real-time detection of print offset:

[0196] Formula 36:

[0197] Formula 37:

[0198] Formula 38:

[0199] in, These are the outline points (pixel coordinates) scanned in real time during the printing process. These are the reference contour points (pixel coordinates) after cutting. X / Y axis deviation (unit: pixels). Set the overall deviation (unit: pixels); set the threshold. ,when It needs to be revised from time to time.

[0200] PID dynamic correction formula, adjusts the print path in real time, and compensates for offset:

[0201] Formula 39:

[0202] Formula 40:

[0203] Among them, compensation offset. X / Y axis correction amount (unit: pixels); This is the proportional coefficient (rapid response deviation). This is the integral coefficient (to eliminate cumulative bias). The differential coefficient is used to suppress overshoot and avoid excessive correction; the adjustment period is 0.3s (synchronized with the camera scanning period) to ensure timely correction.

[0204] The printing parameters are adaptive to suit the printing needs of different materials, ensuring print quality.

[0205] Formula 41:

[0206] Formula 42:

[0207] in, The distance between the nozzle and the material (unit: mm). =1.5~3mm (1.5mm for flexible materials to avoid scratching; 3mm for rigid materials to ensure uniform ink atomization). To print grayscale values ​​(0-255). =1.0~1.2 is the ink grayscale coefficient (1.0 for paper, 1.2 for metal, to adapt to the ink absorption of different materials). Set the grayscale value for the template.

[0208] Step 8: After the operation is completed, the crossbeam is located at the rear of the adsorption platform. The crossbeam 8 moves forward, and the integrated camera 5 scans the finished product. The scanned image is transmitted back to the computer. The computer compares the finished product image with the preset finished product template in grayscale and outline. If the finished product is found to have deviations that exceed the qualified threshold, such as pattern offset, missing print, or cutting defects, the computer marks the specific location of the unqualified finished product on the display screen to complete the automated quality inspection.

[0209] Accurately identify finished product defects (missing prints, cutting flaws, misaligned prints) to achieve markless quality inspection.

[0210] Grayscale comparison (detecting printing defects / color deviations), comparing the color uniformity of the finished product with the standard template:

[0211] Formula 43:

[0212] in, This refers to the grayscale value of the ROI area (Region of Interest, such as a logo or text area) in the standard template. This represents the grayscale value of the ROI region corresponding to the finished product. The pixel size of the ROI region; This represents the mean deviation of grayscale values ​​(unit: grayscale level). A score of ≤10 indicates a passing grade (no missing prints, uniform color); otherwise, it indicates missing prints or uneven color.

[0213] Contour comparison (Hausdorff distance, for detecting cutting defects): Calculate the maximum deviation between the finished contour and the standard contour to determine if the cutting is acceptable.

[0214] Formula 44:

[0215] Where P is the finished contour point set, and Q is the standard contour point set; This represents the maximum and minimum distances from the finished contour points to the standard contour. The maximum and minimum distance from the standard contour point to the finished contour; HD is the Hausdorff distance (unit: mm). HD≤0.2mm, the cutting is qualified; 0.2mm<HD≤0.5mm, slight defects (such as edge burrs), mark the position; HD>0.5mm, serious defects (such as missing contours, over-cutting), directly reject.

[0216] Specialized inspection for printing misalignment: Determines whether the relative position of the printed pattern and the cutting contour is acceptable, and extracts the edge contour of the printed pattern. With cutting contour Calculate the minimum distance between the two:

[0217] Formula 45:

[0218] in, The time is judged as qualified (no misalignment); (The pattern extends beyond the cutting area) or When the distance between the pattern and the cutting edge is too large, it is judged as a misalignment of the cutting.

[0219] Step 9: After quality inspection, manually control the crossbeam 8 to move to the rear standby position of the adsorption platform 7; remove the unqualified products marked on the display screen and take out the qualified finished products; finally, clean the surface of the adsorption platform to remove residual cutting residue and ink marks, ensuring that the platform surface is clean and ready for the next batch of processing.

[0220] In this embodiment, if the system selects the cutting mode and activates the adsorption function, the material is placed on the material platform. The ultrasonic sensor automatically detects the material thickness, and the data is transmitted to the computer in real time. The computer synchronously adjusts the height of the integrated camera 5 (driven by the lifting device 4 to ensure that the camera focal length is adapted to the material thickness) and the cutting depth parameter of the cutting device 3. The integrated camera starts and completes scanning. The computer can either call the preset standard cutting template to accurately align and match it with the actual position of the material (automatically compensating for placement offset), or manually draw a custom cutting path in the control software, or import an externally designed path file. After confirming the cutting scheme, the cutting device 3 starts and completes the cutting operation.

[0221] In this embodiment, if the system selects the printing mode, the material is placed on the adsorption platform, the ultrasonic sensor automatically detects the material thickness, and the data is transmitted to the computer in real time. The computer synchronously adjusts the height of the integrated camera 5, and on the other hand, adjusts the vertical height of the UV printer head 2. The integrated camera 5 starts and completes scanning. The computer calls the preset printing template (such as pattern, color parameters, printing accuracy), and based on the material position obtained by scanning, the printing template is precisely aligned with the material, and the X / Y axis offset is automatically corrected. The UV printer head starts and the printing job is completed.

[0222] In summary, the multi-position intelligent cutting and printing device of this invention can achieve high-precision alignment control throughout the entire process. By acquiring and recognizing images four times—before cutting, after cutting, after the first print, and after printing—it achieves multi-position intelligent visual positioning throughout the entire process, thus constructing a closed-loop processing flow of "cutting—scanning—printing". This technology breaks through the reliance on manual marking and fixed fixtures in traditional cutting and printing processes, significantly improving alignment accuracy and flexible processing capabilities, and is especially suitable for scenarios with irregular materials or large batch deviations.

[0223] The multi-positioning intelligent cutting and printing device of the present invention can improve the robustness of matching the printed pattern with the cutting contour. By integrating the contour matching method of Hu moment and curvature features, and introducing the unified similarity index Sim defined by Formula 27, the device combines the thin plate spline (TPS) and iterative nearest point (ICP) algorithm to achieve global and local path adaptive correction of cutting error, effectively avoiding the risk of misalignment between the pattern and the actual contour, and significantly improving the robustness and adaptability of printing alignment.

[0224] The cutting and printing device with multiple intelligent positioning of the present invention can support dynamic path compensation, improve real-time performance and production line adaptability. During the printing process, the UV print head path is dynamically compensated by real-time image acquisition and deviation detection algorithm (formulas 36-38) based on PID control model (formulas 39-40), so that the pattern can still be accurately output even when the material is offset or deformed, meeting the real-time response requirements of high-speed production line operation.

[0225] The multi-positioning intelligent cutting and printing device of the present invention can construct a judgment and rejection mechanism to improve the yield of finished products. By introducing a fourth image acquisition after printing and calculating grayscale deviation and contour position deviation, the system can accurately identify defects such as missing print, misalignment or cutting deviation, and execute the rejection logic of non-conforming products based on the judgment result, thereby realizing closed-loop control of finished product quality and significantly improving the overall yield.

[0226] The multi-position intelligent cutting and printing device of the present invention can support decoupled operation of cutting mode and printing mode, improve system adaptability and flexibility, and provide three switchable operation modes: cutting mode, printing mode and cutting and printing mode. The execution process and control logic can be flexibly selected according to actual production needs, supporting pure cutting or pure printing process, realizing modular decoupling and multi-task compatibility of the system, and broadening the applicability of the equipment in different processing scenarios.

[0227] The above are merely preferred embodiments of the present invention and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should recognize that any equivalent substitutions and obvious changes made based on the description and illustrations of the present invention should be included within the protection scope of the present invention.

Claims

1. A cutting and printing method with multiple intelligent positioning, characterized in that, The method includes: Step 1: Start the equipment and complete the self-test of the cutting device, lifting device, integrated camera, cleaning device, UV printer head and adsorption platform. Step 2: Select the cutting mode, printing mode, or cutting and printing mode according to the processing task, and initialize the corresponding control parameters; Step 3: Start the negative pressure adsorption of the adsorption platform to fix the material on the adsorption platform, and adjust the imaging height of the integrated camera by the lifting device according to the ultrasonic thickness measurement results; Step 4: Control the crossbeam to drive the integrated camera to perform the first scan, acquire the initial image of the material, and transmit it to the computer system; Step 5: Extract material contours based on the initial image, combine them with a preset cutting template to complete feature matching, generate a cutting path, and drive the cutting device to complete the cutting; Step 6: After cutting, perform cleaning and a second scan to obtain the actual contour image, and use this image to dynamically match and align the printing template. Step 7: Drive the UV printer head to print the pattern along the corrected path. During this process, a third scan is performed and an image is acquired to detect material offset. If the deviation exceeds the threshold, the printing path parameters are adjusted to achieve dynamic compensation for printing alignment. Step 8: After printing is completed, perform a fourth scan and acquire images. Based on grayscale deviation and contour position deviation, identify and mark missing prints, misalignments, or cutting defects. Step 9: Remove defective products according to the defect markings and clean up the platform residue to proceed to the next processing cycle.

2. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, In step 5, the similarity calculation between the printed template outline and the actual outline after cutting is performed using the following formula; ； in: for Moment weights; Curvature weight; For materials Rectangle; template Rectangle; for Euclidean distance; For the curvature of the material; Template curvature; The curvature Euclidean distance; for The maximum possible distance of the moment; The maximum possible distance of curvature; The Hu moment feature distance and curvature feature distance are fused according to a weighted coefficient to obtain a unified similarity index Sim. When Sim is lower than the preset matching threshold, the system triggers local deformation correction based on thin plate spline interpolation and global path alignment based on ICP registration, thereby realizing adaptive adjustment of the printing path to the cutting contour.

3. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, In step 7; The following formula enables real-time detection of X-axis deviation ΔX, Y-axis deviation ΔY, and overall deviation Δ during the printing process; ； in: These are the outline points scanned in real time during the printing process; These are the reference contour points after cutting; for Shaft misalignment; For comprehensive deviation; The dynamic path compensation operation for the above deviation is achieved through the following formula; ； in: for Shaft correction amount; This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; The compensation amount is calculated based on the PID control algorithm and includes three control parameters: proportional response, integral adjustment, and derivative correction. It is used to correct the motion path of the UV printer head in real time to ensure that the printed pattern is aligned with the cutting contour with high precision under unmarked conditions.

4. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, In step 6, the degree of spatial deviation between the printed template outline and the actual outline after cutting is determined by the following formula; ； in: These are the outline points for the printing template; These are the corresponding points of the cut contour; It is the sum of squared differences; The SSD squared difference error calculation operation shown in the figure triggers the recalibration logic of the printing path when the SSD error value exceeds the set tolerance threshold, so as to improve the alignment accuracy and avoid pattern misalignment.

5. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, In step 5, the central moments are normalized to a fixed range using the following formula; ； in: It is the zeroth moment, equal to the pixel area S of the contour; This is a first-order moment used to calculate the centroid of the profile; It is a second moment, describing the degree of dispersion of the profile; ； in: Center distance; The normalized center distance; The coordinates of the centroid of the contour; These are the coordinates of the pixel relative to the centroid; By dividing by The power of the value ensures that the feature value remains unchanged after the material is scaled. The following formulas are used to extract the Hu moment feature and curvature feature of the material profile; ； in: It has 7 Hu units; ； in: For three consecutive sampling points; Let be the curvature of the i-th sampling point; The feature values ​​are used to jointly describe the global shape and local variation features of the target graphic. These feature values ​​serve as the basic input for subsequent template matching and path planning, effectively improving the robustness of matching and alignment stability in an unmarked environment.

6. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, When the control system is set to cutting mode, the method only executes steps 1 to 5, without starting the cleaning device, image correction, or printing. After completing the multi-axis linkage cutting operation of image acquisition, template matching, and target contour, the current processing cycle ends.

7. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, When the control system is set to printing mode, the method only executes steps 1, 2, 3 and 7, and directly starts the pattern printing operation based on the positioning relationship between the material outline and the printing template, omitting the outline cutting, cleaning and secondary scanning steps, which is suitable for finished product processing tasks that do not require structural cutting.

8. The cutting and printing method with multiple intelligent positioning according to claim 1, characterized in that, The method is applied to a cutting and printing device with multiple intelligent positioning, and the cutting and printing device includes: The control system is used to uniformly schedule the various execution modules and run image processing and path planning algorithms. The adsorption platform has a controllable negative pressure adsorption function and is used to fix the material to be processed. An integrated camera is used to acquire original images and cut images before and after processing to achieve multi-stage image recognition and position correction; A lifting device is used to adjust the imaging height of the integrated camera according to the thickness of the material; The cutting device is installed on a three-axis linkage platform and performs material cutting according to the path generated by the control system. The UV printer head is configured to print patterns on materials along a preset path and supports image feedback control. A cleaning device is used to remove debris after cutting, ensuring image quality for both image acquisition and printing; among which... The control system is configured to support the following functional processes: After the device is initialized, the first image acquisition and contour extraction of the material are completed. The matching algorithm is executed to generate the cutting path and drive the cutting device to complete the cutting operation. Then, the image is acquired again to correct the printing template and the deviation detection and path dynamic compensation control are executed to finally achieve the consistency of the cutting-printing closed loop. Furthermore, it supports coordinated scheduling and closed-loop data control of each stage in the processing flow, including image acquisition, contour extraction, path generation, image correction, and deviation compensation, to ensure that the device achieves high-precision consistency between the cutting path and the printed pattern under unmarked conditions.

9. The cutting and printing method with multiple intelligent positioning according to claim 8, characterized in that, The apparatus is suitable for performing only image-guided cutting processes and includes: An integrated camera for capturing raw images of materials; A lifting device is used to adjust the imaging height of the camera; The cutting device, installed on a three-axis linkage platform, is used to perform path cutting after image recognition; The control module is used to extract image contours, perform template matching and path planning, and control the cutting device to perform precise cropping; wherein... The control module supports image alignment and path generation based on contour recognition and feature matching under unmarked conditions, realizing image-execution closed-loop control in pure cutting mode.

10. The cutting and printing method with multiple intelligent positioning according to claim 8, characterized in that, The apparatus is suitable for performing only a pattern printing process and includes: An integrated camera for capturing images of positioned materials; A UV printer head used to print patterns along a set path; The control module is used to compare the image with the print template and correct the print path; among which... The control module supports dynamic correction of the printing path based on SSD error and position offset calculation results, and performs continuous image acquisition and feedback compensation during the printing process to achieve accurate pattern output in pure printing mode.

Images