An Image Processing Method and Device for Coaxial Vision Detection of a Laser Cutting Head

Abstract

A method and device for coaxial vision detection image processing of a laser cutting head. The method includes a static processing method, fixing the laser cutting head and the workpiece, dotting to form a static grayscale image, selecting the bright spot and the nozzle bright ring by framing, extracting the center point of the first bright spot and the center point of the first contour of the nozzle, adjusting the cutting head until the bright spot is coaxial with the nozzle, and generating a reference circle frame area and a ring frame area; a dynamic processing method, including collecting a dynamic grayscale image, real-time detecting the center point of the second bright spot and the center point of the second contour, and ensuring the coaxiality during the dynamic cutting process; the device is composed of an imaging module, a filter module, a beam combining module, a focusing module and a host computer, each module is connected in sequence, and the host computer is electrically connected to the camera to realize functions of image acquisition, processing and cutting head adjustment.

Description

Technical Field

[0001] The present invention relates to the field of laser cutting, and in particular to a method and device for processing coaxial visual detection images of a laser cutting head. Background Art

[0002] Laser cutting technology, especially continuous laser cutting with fiber output, has developed to a power level of tens of kilowatts. In order to achieve effective interaction between laser and material, the design of the laser cutting head is crucial, among which the coaxiality between the nozzle and the bright spot directly affects the uniformity and stability of the cutting. If the bright spot deviates from the center of the nozzle, it will not only lead to a decrease in cutting quality, but also may cause damage to the equipment, especially when using a small nozzle for processing.

[0003] During the laser cutting process, the beam guide module of the laser cutting head may be offset due to high-speed processing and component matching problems. For example, a slight shake of the fiber interface or the rebound gap of the XY centering module may cause the bright spot to shift. Therefore, it is particularly important to detect the coaxiality of the bright spot and the nozzle in real time. In addition, in high-power laser processing, the thermal lens effect will cause the focus of the optical lens to shift, affecting the cutting stability, so real-time monitoring of the bright spot size is also necessary.

[0004] At present, the traditional mechanical vision inspection technology is widely used in the market for bright spot and coaxiality inspection. This method usually completes image processing based on post-processing, and the steps include image acquisition, pre-processing, segmentation and feature extraction. This method has significant shortcomings in processing speed and cannot meet the needs of high-speed laser cutting. Summary of the invention

[0005] The purpose of the present invention is to provide a real-time detection method for the coaxiality of the bright spot of a laser cutting head with the nozzle and the size of the bright spot, which greatly improves the detection efficiency through single-pixel processing function and area segmentation processing technology.

[0006] In a first aspect, the present invention provides a method for processing images of a coaxial visual inspection of a laser cutting head, comprising the steps of:

[0007] A method for processing images of coaxial visual inspection of a laser cutting head comprises S1: forming a static grayscale image of the laser cutting head on a workpiece surface, the static grayscale image comprising a laser bright spot and a nozzle bright ring, and respectively identifying and extracting a bright spot center point of the laser bright spot as a first bright spot center point and a contour center point of the nozzle bright ring as a first contour center point;

[0008] S2: If the center point of the first bright spot coincides with the center point of the first contour, form a reference circular frame area based on the center point of the first bright spot, and form a reference annular frame area based on the center point of the first contour; if the center point of the first bright spot does not coincide with the center point of the first contour, adjust the laser cutting head until the center point of the first bright spot coincides with the center point of the first contour.

[0009] S3: The laser cutting head cuts a real-time acquisition image on the surface of the workpiece to form a dynamic grayscale image, and respectively identifies and extracts the center point of the bright spot in the reference circular frame area in the dynamic grayscale image as the second bright spot center point, and the center point of the contour in the reference annular frame area as the second contour center point. If the second bright spot center point coincides with the second contour center point, it is determined that the bright spot of the laser cutting head is coaxial with the nozzle.

[0010] Further, step S1 of identifying and extracting the center point of the laser bright spot and the center point of the contour of the nozzle bright ring includes the steps:

[0011] S11: Frame the laser bright spot and the nozzle bright ring respectively.

[0012] S12: Based on the framed area of the laser bright spot, extract the center point of the first bright spot.

[0013] S13: Based on the framed area of the nozzle bright ring, extract the center point of the first contour.

[0014] Further, step S12 includes:

[0015] Step S121: For the framed area of the laser bright spot, if the area occupied by the inner edge pixel points in the framed area is greater than 1 / 2, include it in the framed area for gray value comparison.

[0016] Step S122: Extract the pixel point with the maximum gray value as the center point of the first bright spot.

[0017] Further, step S13 includes:

[0018] Step S131: For the framed area of the nozzle bright ring, if the area occupied by the inner edge pixel points in the framed area is greater than 1 / 2, include it in the framed area for gray value comparison.

[0019] Step S132: Compare the gray values of the pixel points row by row or column by column in the framed area of the nozzle bright ring. Take the average value of the inner circle pixel point coordinates when the gray value jump between two adjacent pixel points satisfies the first threshold as the nozzle contour distribution point, and calculate the center point of the first contour based on the contour distribution point.

[0020] Further, in step S132: The first threshold is t times that the high gray value of two adjacent pixel points is the low gray value, where t ≥ 1.5.

[0021] Further: In step S132, if there are multiple groups of adjacent two-pixel point gray value jumps in the same row or the same column that satisfy the first threshold, compare the average value of the coordinates of the multiple groups of adjacent pixel points with the nozzle profile distribution points in the previous and subsequent f rows or columns, and select the average value of the adjacent two-pixel coordinates closest to the nozzle profile distribution points in the previous and subsequent f rows or columns as the nozzle profile distribution points.

[0022] Further, step S2 specifically includes: generating a reference circular frame area with a diameter D based on the center point of the first bright spot, generating a reference annular frame area with a radius R based on the center point of the first contour, obtaining and storing the pixel point data of the reference circular frame area and the reference annular frame area, and uploading the pixel gray values outside the area after resetting them to zero.

[0023] Further, step S3 includes:

[0024] S31: Detect the gray values of all pixel points within the reference circular frame. When the ratio of the gray values of a pixel point and its adjacent P pixel points to the maximum pixel point gray value within the reference circular frame is (K + Δt)%, calculate the relative radius or the equivalent number of pixel points of the relative bright spot area; if no pixel point with a gray value below (K + Δt)% of the maximum gray value can be obtained, repeat step S2;

[0025] S32: Obtain the contour distribution points within the reference annular frame area. If the complete contour distribution points cannot be obtained, repeat step S2.

[0026] Further, calculating the first contour center point based on the contour distribution points in step S132 is based on the best circular arc segment of the contour distribution points as the selection basis.

[0027] Further, in step S2: generating a reference circular frame area with a diameter D based on the center point of the first bright spot, such that at least C pixel points on the boundary of the reference circular frame area have gray values that are (n + Δt)% of the gray value of the center point of the first bright spot;

[0028] Generating a reference annular frame area with a radius R based on the center point of the first contour, such that the number of pixel points in the inner diameter row or column direction of the reference annular frame area is greater than m.

[0029] On the other hand, a coaxial vision detection image processing device for a laser cutting head is provided, including: an imaging module, a filter module, a beam combining module, a focusing module, and a host computer; the imaging module, the filter module, the beam combining module, and the focusing module are assembled together in sequence through an adapter or connection structure, and the host computer is electrically connected to the imaging module camera.

[0030] Further, the imaging module includes a camera and an imaging lens, the filter module includes a band-pass filter and an attenuation sheet, the beam combining module mainly includes a beam combiner, and the focusing module includes a focusing lens, a protective lens, and a fluid structure component including a nozzle. The camera of the imaging module is coaxially arranged with the central axis of the imaging lens, the central axes of the band-pass filter and the attenuation sheet of the filter module are coaxially arranged, and the central axes of the focusing lens and the protective lens of the focusing module are coaxially arranged with the central axis of the beam combiner of the beam combining module.

[0031] Further, a plane mirror is added between the imaging module, the filter module, and the beam combining module for structural transition.

[0032] Further, the band filtering range of the band-pass filter includes at least all bands except the passing band between 300 nm and 1100 nm.

[0033] Further, a gas flow channel structure is arranged inside the nozzle of the focusing module to provide auxiliary air flow to prevent the soot generated during the cutting process from affecting the beam transmission.

[0034] The beneficial effects of the embodiments of the present invention are as follows: A static grayscale image is formed by the laser cutting head in a static state. The center point of the first bright spot and the center point of the first contour of the static grayscale image are identified through image processing, and a reference circular frame area and a reference ring frame area are formed. When the laser cutting head performs dynamic cutting, the cutting image is collected in real time to form a dynamic grayscale image. The center point of the second bright spot and the center point of the second contour of the dynamic grayscale image are identified through image processing. When the center point of the second bright spot coincides with the center point of the second contour during dynamic cutting, it indicates that the bright spot of the laser cutting head is coaxial with the nozzle during dynamic cutting. When the coaxial determination condition is met, the reference circular frame area and the reference ring frame area remain unchanged during dynamic cutting. Only when the coaxial determination condition is not met, the reference circular frame area and the reference ring frame area are dynamically adjusted, reducing the processing time of the system and improving the detection efficiency and detection accuracy. Description of the Drawings

[0035] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0036] Figure 1 It is a schematic diagram of the coaxial vision detection image processing device for the laser cutting head provided by the embodiments of the present invention;

[0037] Figure 2 It is a schematic diagram of the static grayscale processing image provided by the embodiments of the present invention.

[0038] Icons: 1 - Imaging module; 2 - Filter module; 3 - Beam combining module; 4 - Focusing module; 5 - Host computer; 11 - Camera; 12 - Imaging lens; 21 - Band - pass filter; 22 - Attenuator; 31 - Beam combiner; 41 - Focusing lens; 42 - Protective lens; 43 - Nozzle; 61 - Bright spot; 62 - Nozzle bright ring; 63 - Stray light area; 64 - Dark area. Specific embodiments

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Components of the embodiments of the present invention usually described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations.

[0040] Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0041] It should be noted that: Similar reference numerals and letters indicate similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0042] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed during use. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, the terms "first", "second", "third", etc. are only used for descriptive distinction and cannot be understood as indicating or implying relative importance.

[0043] In addition, terms such as "horizontal" and "vertical" do not mean that the components are required to be absolutely horizontal or hanging vertically, but can be slightly inclined. For example, "horizontal" only means that its direction is more horizontal relative to "vertical", and does not mean that the structure must be completely horizontal, but can be slightly inclined.

[0044] In the description of the present invention, it should also be noted that unless otherwise clearly specified and defined, the terms "set", "installed", "connected", and "coupled" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0045] As Figure 1 shown, the present invention provides a method for processing an image of coaxial vision detection of a laser cutting head, including the steps of:

[0046] S1: Form a static grayscale image of the laser cutting head on the surface of the workpiece. The static grayscale image includes a laser bright spot and a nozzle bright ring. Respectively identify and extract the center point of the bright spot of the laser bright spot as the first bright spot center point, and the contour center point of the nozzle bright ring as the first contour center point;

[0047] In step S1, the laser cutting head and the workpiece are fixed and stationary. Under static conditions, the laser cutting head makes a dot on the surface of the workpiece, and the dotting power is the same as the power of the laser cutting head during operation. The imaging module receives the optical signal and converts it into a static grayscale image. The static grayscale image is as Figure 2 shown, including: a bright spot 61, a nozzle bright ring 62, a stray light area 63, and a dark area 64.

[0048] The steps of identifying and extracting the center point of the bright spot of the laser bright spot and the contour center point of the nozzle bright ring in step S1 include the steps of:

[0049] S11: Respectively frame the laser bright spot and the nozzle bright ring;

[0050] S12: Based on the framed area of the laser bright spot, extract the first bright spot center point;

[0051] S13: Based on the framed area of the nozzle bright ring, extract the first contour center point.

[0052] In the above steps, framing the bright spot and the nozzle bright ring can reduce the calculation amount of identifying and extracting the first bright spot center point and the first contour center point, thereby improving the processing efficiency. The first bright spot center point and the first contour center point are the basis for judging whether the bright spot formed by the laser cutting head and the nozzle are coaxial under static conditions.

[0053] Step S12 includes:

[0054] Step S121: For the framed area of the laser bright spot, if the area occupied by the inner edge pixel points in the framed area is greater than 1 / 2, it is included in the framed area for gray value comparison;

[0055] Step S122: Extract the pixel point with the maximum grayscale value as the center point of the first bright spot.

[0056] If there are multiple pixel points with the maximum grayscale value, calculate the average value of the coordinates of the multiple pixel points as the center point of the first bright spot.

[0057] By comparing grayscale values and selecting the pixel point with the maximum grayscale value as the center point of the bright spot, noise interference can be effectively excluded, and the accuracy of extracting the center point of the bright spot can be improved.

[0058] Step S13 includes: Step S131: The framed area of the nozzle bright ring, and if the area occupied by the inner edge pixel points in the framed area is greater than 1 / 2, it is included in the framed area for grayscale value comparison;

[0059] Step S132: Compare the grayscale values of the pixel points row by row or column by column in the framed area of the nozzle bright ring, and use the average value of the coordinates of the inner ring pixel points when the grayscale value jump between two adjacent pixel points satisfies the first threshold as the nozzle contour distribution point, and calculate the first contour center point based on the contour distribution point.

[0060] By comparing grayscale values row by row or column by column and selecting the pixel points that satisfy the first threshold of grayscale value jump as the nozzle contour distribution points and calculating the contour center point, the extraction accuracy of the nozzle contour center point is improved.

[0061] In Step S132: The first threshold is that the high grayscale value of two adjacent pixel points is t times the low grayscale value, where t≥1.5.

[0062] In Step S132: If there are multiple groups of adjacent two pixel points whose grayscale value jumps satisfy the first threshold in the same row or the same column, compare the average values of the coordinates of the multiple groups of adjacent two pixel points with the nozzle contour distribution points in the f rows or columns before and after, and select the average values of the two adjacent pixel coordinates closest to the nozzle contour distribution points in the f rows or columns before and after as the nozzle contour distribution points.

[0063] If there are multiple groups of jump points in the same row or the same column, select the point with the best continuity by combining the data of the previous and next rows, and use the continuity of the nozzle contour to avoid noise, improving the recognition accuracy of the contour distribution points.

[0064] In Step S132, calculating the first contour center point based on the contour distribution point, calculate the first contour center point using the best arc segment of the contour distribution point as the selection basis.

[0065] Using the best arc segment of the contour distribution point to fit and calculate the contour center point can improve the fitting accuracy and thus improve the coaxial detection accuracy.

[0066] S2: If the center point of the first bright spot coincides with the center point of the first contour, form a reference circular frame area based on the center point of the first bright spot, and form a reference ring frame area based on the center point of the first contour; if the center point of the first bright spot does not coincide with the center point of the first contour, adjust the laser cutting head until the center point of the first bright spot coincides with the center point of the first contour.

[0067] If the center point of the first bright spot does not coincide with the center point of the first contour, it is necessary to adjust the centering structure of the laser cutting head to re-punch until the center point of the first bright spot coincides with the center point of the first contour. The coincidence deviation accuracy between the center point of the first bright spot and the center point of the first contour is controlled within the range of 50um, and preferably within the range of 10um. Through the above static punching, that is, statically processing to align the bright spot of the laser cutting head with the nozzle coaxially, ensuring the coaxiality of the bright spot and the nozzle of the laser cutting head before dynamic cutting, and improving the accuracy and stability of laser cutting.

[0068] Step S2 specifically includes:

[0069] Generate a reference circular frame area with a diameter D based on the center point of the first bright spot, generate a reference ring frame area with a radius R based on the center point of the first contour, obtain and store the pixel point data of the reference circular frame area and the reference ring frame area, and upload the pixel gray values outside the area after zeroing.

[0070] Step S2 generates a reference area and stores data, providing a reference area for real-time acquisition and processing during the subsequent dynamic cutting of the laser cutting head. Zeroing the non-reference area can reduce interference and improve the accuracy of data processing.

[0071] In step S2: Generate a reference circular frame area with a diameter D based on the center point of the first bright spot, ensuring that at least C pixel points on the boundary of the reference circular frame area have gray values that are (n + Δt)% of the gray value of the center point of the first bright spot;

[0072] Generate a reference ring frame area with a radius R based on the center point of the first contour, ensuring that the number of pixel points in the inner diameter row or column direction of the reference ring frame area is greater than m.

[0073] Generate a reference circular frame area with a diameter D based on the center point of the bright spot, ensuring that at least C pixel points on the boundary of the reference circular frame area have gray values that are (n + ∆t)% of the gray value of the center point of the bright spot, where C ≥ 3, the value range of n is 1 - 30, and the optimal value of n is 10; ∆t is the gray value deviation, and the value range is ±1 - ±5. The above conditions are used to ensure that the boundary of the reference circular frame area matches the boundary of the actual bright spot.

[0074] Generate a reference ring frame area with a radius R based on the center point of the contour, ensuring that the number of pixel points in the radial row or column direction within the reference ring frame area is greater than m, where the value range of m is 5 - 50, and the optimal value of m is 20. The above limitations can ensure that the reference ring frame area contains sufficient nozzle contour information.

[0075] Among them, the reference circle frame area and the reference ring frame area are reference areas defined by the laser cutting head for real-time detection of the coaxiality of the bright spot and the nozzle during actual dynamic cutting, which reduces the calculation amount, improves the detection efficiency, and can more accurately extract the characteristics of the bright spot and the nozzle during the real-time dynamic cutting process, improving the detection accuracy.

[0076] S3: The laser cutting head cuts on the workpiece surface to collect real-time images to form a dynamic grayscale image, and respectively identifies and extracts the bright spot center point in the reference circle frame area of the dynamic grayscale image as the second bright spot center point, and the contour center point in the reference ring frame area as the second contour center point. If the second bright spot center point coincides with the second contour center point, it is determined that the bright spot of the laser cutting head is coaxial with the nozzle.

[0077] During the dynamic cutting process, based on the position deviation between the second bright spot center point and the second nozzle contour center point, the coaxiality of the bright spot and the nozzle can be evaluated in real time.

[0078] Step S3 includes: S31: Detect the grayscale values of all pixel points within the reference circle frame. When there exists a pixel point and the grayscale values of its adjacent P pixel points have a ratio of (K + Δt)% to the maximum pixel point grayscale value within the reference circle frame, calculate the relative radius or equivalent number of pixel points of the relative bright spot area; if no pixel point with a grayscale value below (K + Δt)% of the maximum grayscale value can be obtained, repeat step S2;

[0079] Extract the bright spot center point in the reference circle frame area, detect the grayscale values of all pixel points within the reference circle frame. When there exists a pixel point and the grayscale values of its adjacent P pixel points have a ratio of (K + Δt)% to the maximum pixel point grayscale value within the reference circle frame, calculate the relative radius or equivalent number of pixel points of the relative bright spot area according to the coordinate mean of the adjacent Q pixel points including this pixel point; where P ≥ 2, Q ≥ 3; the value range of K is 31 - 70, and the optimal value is 50; ∆t is the grayscale value deviation, and the value range is ±1 - ±5; compared with step S2, K > n; where the equivalent number of pixel points refers to all the pixel points within the relative bright spot area. If no pixel point with a grayscale value below (K + Δt)% of the maximum grayscale value can be obtained, repeat step S2. Calculating the relative radius or equivalent number of pixel points of the relative bright spot area can detect whether the bright spot of the laser cutting head becomes larger or whether the bright spot shifts during dynamic operation. If the bright spot becomes larger or the bright spot shifts, defocus compensation can be performed during operation, that is, repeat step S2 to maintain the coaxiality of the bright spot and the nozzle of the laser cutting head during the working process.

[0080] S32: Obtain the contour distribution points within the reference ring frame area. If complete contour distribution points cannot be obtained, repeat step S2.

[0081] If the complete contour distribution points cannot be obtained, it indicates that the contour distribution points of the nozzle are deformed. Based on this, the damage condition of the nozzle can be judged, and problems can be discovered in time to reduce the risk of nozzle damage by burning.

[0082] In this case, a static grayscale image is formed by the laser cutting head in a static state. The center point of the first bright spot and the center point of the first contour of the static grayscale image are identified through image processing, and a reference circular frame area and a reference ring frame area are formed. When the laser cutting head performs dynamic cutting, the cutting image is collected in real time to form a dynamic grayscale image. The center point of the second bright spot and the center point of the second contour of the dynamic grayscale image are identified through image processing. When the center point of the second bright spot coincides with the center point of the second contour during dynamic cutting, it indicates that the bright spot of the laser cutting head is coaxial with the nozzle during dynamic cutting. When the coaxial determination condition is met, the reference circular frame area and the reference ring frame area remain unchanged during dynamic cutting. Only when the coaxial determination condition is not met, the reference circular frame area and the reference ring frame area are dynamically adjusted, reducing the processing time of the system and improving the detection efficiency and detection accuracy.

[0083] In another embodiment, the present invention also provides a coaxial vision detection image processing device for a laser cutting head, which can apply a coaxial vision detection image processing method for a laser cutting head in the above embodiment, including: an imaging module 1, a filter module 2, a beam combining module 3, a focusing module 4, and a host computer 5; the imaging module 1, the filter module 2, the beam combining module 3, and the focusing module 4 are sequentially assembled together through an adapter or connection structure, and the host computer 5 is electrically connected to the imaging module 1.

[0084] The imaging module 1 includes a camera 11 and an imaging lens 12. The filter module 2 includes a band-pass filter 21 and an attenuation sheet 22. The beam combining module 3 mainly includes a beam combining mirror 31. The focusing module 4 includes a focusing lens 41, a protective lens 42, and a fluid structure component including a nozzle 43. Among them, the central axes of the camera 11 and the imaging lens 12 of the imaging module 1 are coaxially arranged. The central axes of the band-pass filter 21 and the attenuation sheet 22 of the filter module 2 are coaxially arranged. The central axes of the focusing lens 41 and the protective lens 42 of the focusing module 4 are coaxially arranged with the central axis of the beam combining mirror 31 of the beam combining module 3.

[0085] Among them, camera 11 is a pixel-internal processing type camera. Each pixel point includes three micro-units: a sensor, a memory, and a processor. The memory stores the data obtained by the sensor, and the processor processes the data obtained by the sensor. Each pixel point is coplanar and electrically connected. Each pixel point of the camera has the ability to store and process data, can quickly process image data, and improve the detection efficiency. The imaging lens includes, but is not limited to, a fixed-focus lens and a zoom lens, and at least has the functions of adjusting the light input amount and adjusting the imaging position. Its effective focal length ranges from 10 mm to 150 mm, typically 25 mm, 35 mm, 50 mm, 75 mm, or a continuously adjustable focal length within the range. The imaging lens supports both fixed focus and zoom, has a wide focal length range, and can be adjusted according to different cutting scenarios.

[0086] A flat mirror can be added between the imaging module 1, the filter module 2, and the beam combining module 3 for structural transition.

[0087] The band-pass filtering range of the band-pass filter 21 includes at least all bands except the passing band between 300 nm and 1100 nm. Typically, there are 480 nm, 532 nm, 580 nm, 650 nm, 780 nm, 940 nm, etc. The half-wave width is within 20 nm, typically 10 nm, 5 nm, 3 nm, etc. The attenuation sheet 22 is used in conjunction with the band-pass filter 21 to attenuate the transmittance of the passing band, including fixed OD and gradient OD. OD represents the cut-off depth. The band-pass filter and the attenuation sheet can precisely control the band and intensity of light, ensuring the imaging quality.

[0088] The focusing lens 41 and the protective lens 42 in the focusing module 4 can efficiently transmit the laser and the passing band of the filter module. The focusing lens 41 focuses the laser to the end of the nozzle 43 and is also used for imaging in cooperation with the imaging lens 12 of the imaging module 1.

[0089] The beam combining mirror 31 can transmit or reflect the laser. Correspondingly, it can also reflect or transmit the passing band of the filter module 2.

[0090] A gas flow channel structure is arranged inside the nozzle 43 of the focusing module 4 to provide auxiliary air flow and prevent the soot generated during the cutting process from affecting the beam transmission.

[0091] The host computer 5 is electrically connected to the camera 11 of the imaging module 1 and has the functions of acquiring the data of each pixel point of the camera 11, image processing and display, and issuing processing commands. The host computer integrates control to realize the rapid acquisition, processing, and display of image data, and at the same time has the function of issuing commands, which is convenient for automated operation.

[0092] Along the light beam transmission path, during the process of laser cutting a material, light beams of different wavelengths enter the interior of the laser cutting head through the nozzle 43 opening, successively passing through the protective mirror 42 and the focusing lens 41 of the focusing module 4, and then being reflected or transmitted by the beam combining mirror 31 of the beam combining module 3. The reflected or transmitted light is filtered and attenuated by the band-pass filter 21 and the attenuation sheet 22 of the filter module 2, and the final transmitted wavelength band is focused through the imaging lens 12 of the imaging module 1 and imaged on the camera 11.

[0093] The above coaxial vision detection image processing device for a laser cutting head is applicable to vision detection during the laser cutting process, capable of real-time monitoring of the cutting effect, real-time detection of whether the bright spot 61 and the nozzle 43 are coaxial, and real-time detection of whether the nozzle 43 is deformed, damaged, etc., improving the cutting accuracy and quality.

[0094] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A laser cutting head coaxial visual inspection image processing method, characterized in that: S1: forming a static grayscale image of the laser cutting head on the surface of the workpiece, the static grayscale image including a laser bright spot and a nozzle bright ring, and respectively identifying and extracting a bright spot center point of the laser bright spot as a first bright spot center point and a contour center point of the nozzle bright ring as a first contour center point; S2: If the center point of the first bright spot coincides with the center point of the first contour, a reference circular frame area is formed according to the center point of the first bright spot, and a reference ring frame area is formed according to the center point of the first contour; if the center point of the first bright spot does not coincide with the center point of the first contour, the laser cutting head is adjusted until the center point of the first bright spot coincides with the center point of the first contour; S3: The laser cutting head forms a dynamic grayscale image by cutting the workpiece surface in real time, and respectively identifies and extracts the bright spot center point of the reference circular frame area in the dynamic grayscale image as the second bright spot center point, and the contour center point of the reference ring frame area as the second contour center point. If the second bright spot center point coincides with the second contour center point, it is determined that the bright spot of the laser cutting head is coaxial with the nozzle. Wherein, step S3 comprises: S31: Detect the grayscale values ​​of all pixels in the reference circle frame, and when the ratio of the grayscale value of a pixel and its P neighboring pixels to the grayscale value of the maximum pixel in the reference circle frame is (K+Δt)%, calculate the relative radius of the relative bright spot area or the equivalent number of pixels; if no pixel with a grayscale value below (K+Δt)% of the maximum grayscale value can be obtained, repeat step S2; The value range of K is 31-70, and Δt is the gray value deviation; S32: Obtain the contour distribution points within the reference ring frame area. If the complete contour distribution points cannot be obtained, repeat step S2; If the complete contour distribution points cannot be obtained, it means that there is deformation in the nozzle contour distribution points. Based on this, the damage condition of the nozzle can be judged, and the problem can be discovered in time to reduce the risk of nozzle damage and burning.

2. A laser cutting head coaxial visual inspection image processing method according to claim 1, characterized in that: In step S1, identifying and extracting the bright spot center point of the laser bright spot and the contour center point of the nozzle bright ring includes the following steps: S11: Select the laser bright spot and the nozzle bright ring respectively; S12: extracting the center point of the first bright spot based on the framed area of ​​the laser bright spot; S13: Extracting the first contour center point based on the framed area of ​​the nozzle bright ring.

3. A laser cutting head coaxial visual inspection image processing method according to claim 2, characterized in that: Step S12 includes: Step S121: The framed area of ​​the laser bright spot, the edge pixels in the framed area that occupy an area greater than 1 / 2 are counted into the framed area for grayscale value comparison, Step S122: extracting the pixel with the largest gray value as the center point of the first bright spot.

4. The method for coaxial visual inspection image processing of a laser cutting head according to claim 2, characterized in that: Step S13 includes: Step S131: The framed area of ​​the nozzle bright ring, the edge pixels in the framed area that occupy an area greater than 1 / 2 are counted into the framed area for grayscale value comparison; Step S132: compare the grayscale values ​​of the pixels in the selected area of ​​the nozzle bright ring row by row or column by column, take the average coordinates of the inner circle pixels when the grayscale value jump of two adjacent pixels meets the first threshold as the nozzle contour distribution point, and calculate the first contour center point based on the contour distribution point.

5. The method for coaxial visual inspection image processing of a laser cutting head according to claim 4 is characterized in that: In step S132: the first threshold is t times the high grayscale value of two adjacent pixels as the low grayscale value, where t≥1.

5.

6. A method for coaxial visual inspection image processing of a laser cutting head according to claim 4 or 5, characterized in that: In step S132: if there are multiple groups of adjacent two pixel points in the same row or the same column whose grayscale value jumps meet the first threshold, the coordinate mean of the multiple groups of adjacent two pixel points are compared with the nozzle contour distribution points of the front and back f rows or columns, and the average of the adjacent two pixel coordinate points closest to the contour distribution points of the front and back f rows or columns is selected as the nozzle contour distribution point.

7. The method for coaxial visual inspection image processing of a laser cutting head according to claim 1, characterized in that: Step S2 specifically includes: Generate a reference circular frame area with a diameter of D based on the center point of the first bright spot, and generate a reference ring frame area with a radius of R based on the center point of the first contour, obtain and store pixel point data of the reference circular frame area and the reference ring frame area, and reset the grayscale values ​​of pixels outside the area to zero and upload.

8. The method for coaxial visual inspection image processing of a laser cutting head according to claim 6, characterized in that: The first contour center point is calculated based on the contour distribution points in step S132, with the best arc segment of the contour distribution points being used as a selection basis.

9. A laser cutting head coaxial visual inspection image processing method according to claim 1 or 2, characterized in that: In step S2: Generate a reference circular frame area with a diameter D based on the first bright spot center point, satisfying that the grayscale value of at least C pixels on the boundary of the reference circular frame area is (n+Δt)% of the grayscale value of the first bright spot center point, where n is in the range of 1-30 and Δt is the grayscale value deviation; A reference ring frame area with a radius R is generated according to the first contour center point, and the number of pixel points in the row or column direction in the reference ring frame area is greater than m, and the value range of m is 5-50.

10. A laser cutting head coaxial visual detection image processing device, applied to a laser cutting head coaxial visual detection image processing method according to any one of claims 1 to 9, characterized in that: include: The imaging module, the filtering module, the beam combining module, the focusing module and the host computer are composed; the imaging module, the filtering module, the beam combining module and the focusing module are assembled together in sequence through a connecting structure, and the host computer is electrically connected to the imaging module camera.

11. A laser cutting head coaxial visual inspection image processing device according to claim 10, characterized in that: The imaging module includes a camera and an imaging lens, the filtering module includes a bandpass filter and an attenuation plate, the beam combining module includes a beam combining mirror, and the focusing module includes a focusing mirror, a protective mirror, and a fluid structure component including a nozzle, wherein the camera of the imaging module is coaxially arranged with the central axis of the imaging lens, the bandpass filter and the attenuation plate of the filtering module are coaxially arranged with the central axis, and the focusing mirror and the protective mirror of the focusing module are coaxially arranged with the central axis of the beam combining mirror of the beam combining module.

12. The laser cutting head coaxial visual inspection image processing device according to claim 10, characterized in that: A flat mirror is added between the imaging module, the filter module and the beam combining module to perform structural transition.

13. The laser cutting head coaxial visual inspection image processing device according to claim 11, characterized in that: The band-pass filter has a wavelength-removal range of at least including all wavelengths between 300 nm and 1100 nm except the passing wavelength.

14. The laser cutting head coaxial visual inspection image processing device according to claim 10, characterized in that: A gas flow channel structure is arranged inside the nozzle of the focusing module to provide auxiliary airflow to prevent smoke and dust generated during the cutting process from affecting the light beam transmission.

Images