Visual intelligent identification cutting method, computer equipment and readable storage medium

Abstract

The invention discloses a visual intelligent identification cutting method, computer equipment and a readable storage medium, and relates to the technical field of cutting, the method comprises the following steps: obtaining a breadth image on a workbench of a cutting machine tool, and laying a to-be-cut piece on the workbench of the cutting machine tool; the breadth image is preprocessed; matching a preset reference cutting piece template with the pre-processed breadth image so as to identify all target cutting piece images from the breadth image; according to the angle information and the position information of the target cut-part image, the real-time angle and the real-time position of a preset CAD graph corresponding to the target cut-part image are adjusted, so that the real-time angle and the real-time position of the CAD graph are consistent with the real-time angle and the real-time position of the corresponding target cut-part image; constructing a line path according to all the adjusted CAD graphs; and the cutting machine tool is driven to cut the to-be-cut piece according to the line path. According to the edge cutting device, edge cutting can be accurately achieved, errors caused by manual edge cutting are reduced, and more standardization is achieved.

Description

Technical Field

[0001] The present invention relates to the field of cutting technology, and in particular to a visual intelligent recognition cutting method, a computer device and a readable storage medium. Background Art

[0002] With the continuous improvement of production technology, the quality requirements for blade cutting are inevitably improved. At the same time, with the continuous development of computer vision technology, computer vision has also been applied to many fields. Therefore, it is of great significance to use image recognition technology in computer vision to achieve efficient and rapid recognition in the blade cutting industry.

[0003] Currently, the combination of image recognition and blade cutting technology is a common method for cutting flexible materials (such as fabric and leather). However, this method still has drawbacks: existing blade cutting equipment requires manual laying of the cut pieces, and each time the pieces are laid, there will be position and / or angle deviations, resulting in misalignment between the intended cutting line and the actual cutting line. Furthermore, if workers make temporary adjustments to the flexible material, not only is there no guarantee of a perfect fit, but there are also certain operational safety risks.

[0004] Therefore, it is necessary to develop a visual intelligent recognition cutting method to solve the above problems. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a visual intelligent recognition cutting method, computer equipment and readable storage medium, which can accurately achieve trimming and reduce the errors caused by manual trimming.

[0006] In order to solve the above technical problems, the present invention provides a visual intelligent recognition and cutting method, including: obtaining a format image on a workbench of a cutting machine tool, on which a piece to be cut is laid; preprocessing the format image; matching a preset reference piece template with the preprocessed format image to identify all target piece images from the format image; adjusting the real-time angle and real-time position of a preset CAD graphic corresponding to the target piece image according to the angle information and position information of the target piece image, so that the real-time angle and real-time position of the CAD graphic are consistent with the real-time angle and real-time position of the corresponding target piece image; constructing a line path according to all adjusted CAD graphics; and driving the cutting machine tool to cut the piece to be cut according to the line path.

[0007] As an improvement to the above scheme, the steps of constructing the reference template include: obtaining a reference piece image of the piece to be cut; preprocessing the reference piece image; performing edge detection on the piece to be cut in the preprocessed reference piece image to identify the graphic contour of the piece to be cut; and extracting feature points of the graphic contour to construct a reference piece template of the piece to be cut.

[0008] As an improvement to the above scheme, the preprocessing steps include: performing color conversion processing on the target image, the target image including a format image or a reference cut piece image; performing grayscale processing on the target image after color processing; performing binarization processing on the target image after grayscale processing; and performing blurring processing on the target image after binarization processing.

[0009] As an improvement to the above scheme, the step of performing edge detection on the piece to be cut in the preprocessed reference piece image to identify the graphic outline of the piece to be cut includes: detecting the piece edge of the piece to be cut by an edge detection method; optimizing the piece edge by a quadratic function interpolation method of a gradient modulus; and correcting the optimized piece edge in the horizontal and vertical directions respectively by a median interpolation method.

[0010] As an improvement to the above-mentioned scheme, the step of extracting feature points of the graphic contour to construct a reference cutting piece template for the cutting piece to be cut includes: smoothing the graphic contour by an image pyramid down-sampling method to reduce the resolution of the graphic contour; extracting feature points from the smoothed graphic contour, and recording the point position information of the feature points according to a preset rotation step size; and integrating the point position information into a reference cutting piece template.

[0011] As an improvement to the above scheme, the step of matching a preset reference cutting piece template with the pre-processed format image to identify all target cutting piece images from the format image includes: matching the point information of the reference cutting piece template with the target cutting piece in the format image respectively; when matching, comparing according to a preset rotation step size and a preset proportional step size to calculate the real-time score of each target cutting piece image respectively; when the real-time score of the target cutting piece image reaches the set target score, it indicates that the target cutting piece image is matched successfully.

[0012] As an improvement to the above scheme, the steps of adjusting the real-time angle and real-time position of the preset CAD graphic corresponding to the target cutting piece image according to the angle information and position information of the target cutting piece image include: extracting the preset CAD graphic corresponding to the target cutting piece image from the CAD graphic library; comparing the centers of the minimum circumscribed rectangles of the target cutting piece image and the corresponding preset CAD graphic to adjust the corresponding preset CAD graphic; moving the CAD graphic to the position of the corresponding target cutting piece image; and adjusting the size of the CAD graphic according to a preset shrinkage value.

[0013] As an improvement of the above scheme, the step of driving the cutting machine to cut the piece to be cut according to the line path includes: sorting the line information in the line path; allocating the sorted line information to the cutting beam group of the cutting machine, and the cutting beam group includes at least two cutting beams; constructing a corresponding cutting path according to the line information allocated to each cutting beam; and driving the corresponding cutting beams to perform multi-axis interpolation cutting according to the cutting path.

[0014] Correspondingly, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the steps of the above-mentioned visual intelligent recognition cutting method when executing the computer program.

[0015] Correspondingly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-mentioned visual intelligent recognition and cutting method.

[0016] The implementation of the present invention has the following beneficial effects:

[0017] The present invention combines a reference cutting piece template, a format image and a CAD drawing; by specifically constructing a reference cutting piece template and matching the reference cutting piece template with the actual format image, the target cutting piece image is accurately identified; the target cutting piece image is then combined with the CAD drawing to construct a precise line path; thereby quickly achieving the trimming process that fits the flexible material, accurately achieving trimming, greatly reducing manual intervention, reducing errors caused by manual trimming, saving manpower costs, and achieving greater standardization. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of the first embodiment of the visual intelligent recognition cutting method of the present invention;

[0019] Figure 2 This is a flow chart of the second embodiment of the visual intelligent recognition cutting method of the present invention. DETAILED DESCRIPTION

[0020] To make the objectives, technical solutions, and advantages of the present invention more apparent, the present invention will be further described in detail below with reference to the accompanying drawings. It is hereby stated that any directional terms such as "up," "down," "left," "right," "front," "back," "inside," and "outside" that appear or will appear herein are based solely on the accompanying drawings and are not intended to limit the present invention.

[0021] See also Figure 1 , Figure 1 The flowchart of the first embodiment of the visual intelligent recognition cutting method of the present invention is shown, which includes:

[0022] S101, acquiring a format image on a workbench of a cutting machine;

[0023] It should be noted that pieces to be cut are laid on the workbench of the cutting machine.

[0024] Before work, workers manually place the cutting pieces on the workbench of the cutting machine, and then use a camera to capture a format image of the cutting pieces on the entire workbench.

[0025] S102, pre-processing the format image;

[0026] Furthermore, the steps of pre-processing the format image include:

[0027] (1) Perform color conversion on the format image;

[0028] It should be noted that RGB images correspond to hardware output, while HSV images are more in line with the intuitive vision of the human eye. Therefore, when processing images, RGB images are often converted to HSV images first, and the images are processed in the HSV color space. After processing, the HSV images are converted back to RGB images.

[0029] The present invention converts the RGB channels of an RGB image into HSV (hue, saturation, brightness) channels, and screens out the image portion that needs to be identified by adjusting the HSV (the three parameters have a maximum and minimum range, and the range can be adjusted by a software slider), thereby filtering out interference items outside the required HSV range.

[0030] (2) grayscale processing is performed on the color-processed image;

[0031] It should be noted that color photos are generally three-channel (RGB), and each pixel is composed of three color components. In order to reduce the amount of calculation and interference, the three channels can be processed into a single channel, and the color relationship is converted into a brightness relationship. Each vector is determined by only one component, and 0 to 255 corresponds to dark to bright.

[0032] (3) Binarization processing is performed on the grayscale processed image;

[0033] After grayscale processing, if you still want to filter out suitable graphics and leave only black and white images, you can dynamically adjust the threshold of the image through the slider. For pixels with grayscale values greater than the set threshold, directly change the grayscale value to the maximum (the maximum value can be set by yourself, and the maximum setting range can only be 255), and adjust the grayscale of pixels less than the set threshold to the minimum 0.

[0034] (4) Perform blur processing on the binary image.

[0035] After grayscale processing, the color difference at the edge is already quite obvious, but pixel aliasing and noise still exist. This invention adds Gaussian blur and uses a convolution kernel, scanning each pixel in the image. The weighted average grayscale value of the pixels in the neighborhood determined by the convolution kernel is used to replace the value of the central pixel, achieving a filtering effect, thereby making the identified edges smoother.

[0036] S103, matching a preset reference cutting piece template with the pre-processed format image to identify all target cutting piece images from the format image;

[0037] It should be noted that the reference cutting piece template can be pre-made according to the cutting piece to be cut. By matching the pre-made reference cutting piece template with the format image, the target cutting piece image can be found. The specific steps include:

[0038] (1) Match the point information of the reference cutting piece template with the target cutting piece in the format image respectively;

[0039] (2) During matching, a comparison is performed based on a preset rotation step size and a preset scale step size to calculate the real-time score of each target piece image;

[0040] (3) When the real-time score of the target piece image reaches the set target score, it means that the target piece image is matched successfully.

[0041] That is to say, the point information of the reference cutting piece template is matched with the target cutting piece in the format image, and compared according to the preset rotation step and proportional step. The worker adjusts the target score (i.e., the matching similarity) to obtain a relatively stable value that can match all the target cutting pieces with high accuracy. When the set target score is reached, it will be considered a successful match.

[0042] S104, adjusting the real-time angle and real-time position of a preset CAD graphic corresponding to the target piece image according to the angle information and the position information of the target piece image, so that the real-time angle and the real-time position of the CAD graphic are consistent with the real-time angle and the real-time position of the corresponding target piece image;

[0043] It should be noted that different angle information and position information of the matched target cut piece image can be obtained according to the reference cut piece template.

[0044] Accordingly, the steps of adjusting the real-time angle and real-time position of the preset CAD graphic corresponding to the target piece image according to the angle information and position information of the target piece image include:

[0045] (1) extracting a preset CAD graphic corresponding to the target piece image from the CAD graphic library;

[0046] (2) comparing the center of the minimum circumscribed rectangle of the target piece image and the corresponding preset CAD figure to adjust the corresponding preset CAD figure;

[0047] (3) moving the CAD graphics to the position of the corresponding target piece image;

[0048] (4) Adjust the size of the CAD drawing according to the preset indentation value.

[0049] That is, by adjusting the CAD lines of the CAD drawing according to the different angles and different positions of each matched target piece image (i.e., comparing the centers of the minimum circumscribed rectangles of the target piece image and the corresponding preset CAD drawing to adjust the corresponding preset CAD drawing), and moving the standard CAD lines to the positions of the corresponding target piece images, the CAD drawing can finally be resized according to the set indentation value.

[0050] S105, constructing a line path according to all the adjusted CAD graphics;

[0051] Furthermore, the adjusted CAD graphics can be converted into visual drawing lines and drawn on the format image, allowing workers to see the preview of the cutting line effect; finally, after the workers confirm that it is correct, the line path is sent for cutting.

[0052] S106, driving the cutting machine to cut the piece to be cut according to the line path.

[0053] Furthermore, the step of driving the cutting machine to cut the piece to be cut according to the line path includes:

[0054] (1) Sorting the line information in the line path;

[0055] (2) allocating the sorted line information to the cutting beam group of the cutting machine;

[0056] Wherein, the cutting beam group includes at least two cutting beams;

[0057] (3) Constructing the corresponding cutting path according to the line information assigned to each cutting beam;

[0058] (4) According to the cutting path, the corresponding cutting beams are driven to perform multi-axis interpolation cutting.

[0059] It should be noted that after the line information is sorted, it is distributed to two or more cutting beams in the most efficient cutting method; at the same time, combined with some optimized path processing of corners or V-notches, a smooth cutting process is obtained, thereby controlling the motor to perform multi-axis interpolation cutting, and the cutting process is not stuck and no machine head collision occurs.

[0060] Therefore, the present invention can realize the trimming process of the flexible material at a high speed, reduce manual intervention, and save manpower expenses; at the same time, it can also realize trimming accurately, reduce the error caused by manual trimming, and achieve more standardization.

[0061] See also Figure 2 , Figure 2 A flow chart of a second embodiment of the visual intelligent recognition cutting method of the present invention is shown, which includes:

[0062] S201, obtaining a reference piece image of a piece to be cut;

[0063] Scan / photograph the piece to be cut that requires this function to obtain a reference piece image; when scanning / photographing, the piece to be cut should be placed as close as possible to the angle of the corresponding CAD drawing in the CAD drawing library.

[0064] S202, pre-processing the reference piece image;

[0065] Furthermore, the step of preprocessing the reference piece image includes:

[0066] (1) Performing color conversion on the reference piece image;

[0067] (2) grayscale processing is performed on the color-processed reference piece image;

[0068] (3) Binarization processing is performed on the grayscale processed reference piece image;

[0069] (4) Perform fuzzy processing on the reference piece image after binarization processing.

[0070] The specific pre-processing steps are the same as step S102 in the first embodiment and will not be repeated here. S203, edge detection is performed on the to-be-cut pieces in the pre-processed reference piece image to identify the outline of the to-be-cut pieces;

[0071] Sub-pixel edge detection is performed on the to-be-cut pieces in the pre-processed reference piece image. The specific steps include:

[0072] (1) Detecting the edge of the piece to be cut by edge detection method;

[0073] The edges of the pieces to be cut can be detected by the Canny edge detection method.

[0074] (2) Optimizing the edges of the pieces by using the quadratic function interpolation method of the gradient modulus;

[0075] Since the edges detected by the Canny edge detection method are at the pixel level, the accuracy is low and it is easy to produce jagged edges; the present invention optimizes the edge of the cut piece by calculating the quadratic function interpolation of the gradient modulus values at three adjacent points in the gradient direction, instead of using the non-maximum suppression method in the Canny edge detection method to achieve optimization.

[0076] (3) The optimized edges of the pieces are corrected in the horizontal and vertical directions respectively through the median interpolation method.

[0077] Interpolate the intermediate values of the jittered lines in the horizontal and vertical directions to make them fit more closely, so that the acquired points can reach the sub-pixel level.

[0078] S204: extracting feature points of the graphic outline to construct a reference cutting piece template for the cutting piece to be cut.

[0079] Extract various points from the identified circular contour and integrate the reference cutting template. The specific steps include:

[0080] (1) Smoothing the image contours through image pyramid downsampling to reduce the resolution of the image contours;

[0081] (2) extracting feature points from the smoothed contour of the graphic and rotating and recording the position information of the feature points according to a preset rotation step size;

[0082] (3) Integrate the point information into a reference cutting template.

[0083] That is to say, the pyramid downsampling method is used to continuously reduce the resolution of the image, and then the process of looping the image contour is repeated 7 times. The various identified points are rotated and recorded according to the set rotation step (rotation accuracy). Finally, this series of information (angle range, scale range, number of features, angle step, scale step) is integrated into a template yam l file.

[0084] S205, acquiring a format image on the workbench of the cutting machine;

[0085] The workbench of the cutting machine is provided with pieces to be cut.

[0086] S206, pre-processing the format image;

[0087] S207, matching the preset reference cutting piece template with the pre-processed format image to identify all target cutting piece images from the format image;

[0088] S208, adjusting the real-time angle and real-time position of a preset CAD graphic corresponding to the target piece image according to the angle information and the position information of the target piece image, so that the real-time angle and the real-time position of the CAD graphic are consistent with the real-time angle and the real-time position of the corresponding target piece image;

[0089] S209, constructing a line path based on all adjusted CAD graphics;

[0090] S210: driving a cutting machine tool to cut the piece to be cut according to the line path.

[0091] Therefore, the present invention constructs a reference cutting template in a targeted manner and matches the reference cutting template with the actual format image, thereby accurately identifying the target cutting image; then the target cutting image is combined with the CAD graphics to construct a precise line path, thereby accurately achieving edge cutting, greatly reducing manual intervention and saving manpower expenses.

[0092] Correspondingly, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the steps of the above-mentioned visual intelligent recognition cutting method when executing the computer program.

[0093] At the same time, the present invention also discloses a computer-readable storage medium on which a computer program is stored, wherein when the computer program is executed by a processor, the steps of the above-mentioned visual intelligent recognition and cutting method are implemented.

[0094] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A visual intelligent recognition cutting method, characterized in that: include: Acquire a format image on a workbench of a cutting machine, where pieces to be cut are laid on the workbench of the cutting machine; Preprocessing the image format; Matching a preset reference cutting piece template with the pre-processed format image to identify all target cutting piece images from the format image; adjusting the real-time angle and real-time position of a preset CAD graphic corresponding to the target piece image according to the angle information and the position information of the target piece image, so that the real-time angle and the real-time position of the CAD graphic are consistent with the real-time angle and the real-time position of the corresponding target piece image; Construct line paths based on all adjusted CAD graphics; The cutting machine is driven according to the line path to cut the pieces to be cut.

2. The visual intelligent recognition cutting method according to claim 1, characterized in that: The steps of constructing the reference template include: Acquire a reference piece image of a piece to be cut; Preprocessing the reference piece image; Performing edge detection on the to-be-cut piece in the pre-processed reference piece image to identify a graphic outline of the to-be-cut piece; Feature points of the graphic outline are extracted to construct a reference piece template for the piece to be cut.

3. The visual intelligent recognition cutting method according to claim 1 or 2, characterized in that: The pre-processing steps include: Performing color conversion processing on a target image, wherein the target image includes a format image or a reference piece image; Performing grayscale processing on the target image after color processing; Performing binarization processing on the target image after grayscale processing; The target image after the binarization processing is blurred.

4. The visual intelligent recognition cutting method according to claim 2, characterized in that: The step of performing edge detection on the to-be-cut piece in the pre-processed reference piece image to identify the graphic outline of the to-be-cut piece includes: Detecting the edges of the pieces to be cut by an edge detection method; Optimizing the edge of the cut piece by using a quadratic function interpolation method of a gradient modulus; The optimized edges of the cutting pieces are corrected in the horizontal and vertical directions respectively by using a median interpolation method.

5. The visual intelligent recognition cutting method according to claim 2, characterized in that: The step of extracting the feature points of the graphic outline to construct the reference piece template for the piece to be cut comprises: Smoothing the outline of the graphic by using an image pyramid down-sampling method to reduce the resolution of the outline of the graphic; Extracting feature points from the smoothed graphic contour, and recording the position information of the feature points by rotating them according to a preset rotation step size; The point information is integrated into a reference cutting piece template.

6. The visual intelligent recognition cutting method according to claim 1, characterized in that: The step of matching a preset reference cutting piece template with the pre-processed format image to identify all target cutting piece images from the format image includes: Matching the point information of the reference cutting piece template with the target cutting piece in the format image respectively; During matching, the preset rotation step size and the preset scale step size are compared to calculate the real-time score of each target piece image; When the real-time score of the target piece image reaches the set target score, it means that the target piece image is matched successfully.

7. The visual intelligent recognition cutting method according to claim 1, characterized in that: The step of adjusting the real-time angle and real-time position of the preset CAD graphic corresponding to the target cut piece image according to the angle information and position information of the target cut piece image comprises: Extracting a preset CAD graphic corresponding to the target piece image from a CAD graphic library; Comparing the target piece image with the center of the minimum circumscribed rectangle of the corresponding preset CAD figure to adjust the corresponding preset CAD figure; Moving the CAD graphic to the position of the corresponding target piece image; The size of the CAD graphic is adjusted according to a preset shrinkage value.

8. The visual intelligent recognition cutting method according to claim 1, characterized in that: The step of driving the cutting machine to cut the pieces to be cut according to the line path comprises: sorting the line information in the line path; Distributing the sorted line information to a cutting beam group of a cutting machine tool, wherein the cutting beam group includes at least two cutting beams; According to the line information assigned to each cutting beam, the corresponding cutting path is constructed; According to the cutting path, the corresponding cutting beams are driven respectively to perform multi-axis interpolation cutting.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the visual intelligent recognition and cutting method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the visual intelligent recognition and cutting method according to any one of claims 1 to 8 are implemented.

Images