Molten pool temperature field in-situ monitoring method and system based on multi-view light field imaging

Abstract

The invention provides a molten pool temperature field in-situ monitoring method and system based on multi-view light field imaging, and the method employs a light field camera multi-view system, obtains molten pool images of R, G and B channels through building a light field sub-aperture Bayer model, and calibrates the relative spectral responsivity of a light field camera and the proportion of dual-wavelength intensity under a coaxial temperature measurement system. Molten pool images of two channels are selected from the molten pool images of the R channel, the G channel and the B channel, calibration parameters are combined, and a molten pool temperature field is derived through a dual-wavelength temperature measurement method. A blackbody furnace is adopted to calibrate a multi-view light field temperature measurement system, and the error of a monitored temperature field is kept below 3%. The method simplifies the monitoring process, improves the temperature measurement precision, has a wide temperature range and low complexity, can perform in-situ monitoring on the high-temperature metal molten pool, and has a wide application prospect in the field of additive manufacturing.

Description

Technical Field

[0001] The present invention relates to the field of industrial vision technology, and in particular, to a method and system for in-situ monitoring of the molten pool temperature field based on multi-view light field imaging. Background Art

[0002] Laser metal additive manufacturing refers to a process of continuously melting metal with a high-energy density laser and stacking it layer by layer to prepare metal parts with complex shapes and high density, which has been applied in core fields such as aerospace. Some key process variables in laser metal additive manufacturing, such as the molten pool temperature, can be used to characterize the quality of parts. However, due to the high temperature and fast change of the molten pool, achieving high-precision monitoring of the molten pool temperature field for component performance control is still a key obstacle hindering the wide application of metal additive manufacturing technology. Therefore, the research on the in-situ monitoring technology of the molten pool temperature field during the additive manufacturing process is an important way to solve this problem.

[0003] Currently, the main hardware devices used to monitor the molten pool temperature include commercial devices such as infrared cameras and photodiodes, as well as dual-wavelength temperature measurement systems based on single cameras or dual cameras. Among them, infrared temperature measurement devices need to accurately and repeatedly calibrate the continuously changing material emissivity. However, it is difficult to accurately calibrate the emissivity of materials due to problems such as temperature or surface state changes, especially it is difficult to relate the infrared signal to the accurate temperature during the phase change process. In addition, for some powders, such as Ti6AI4V alloy, the molten pool temperature can exceed 3000°C, which greatly limits the applicability of commercial devices. Although the dual-wavelength high-temperature measurement method has been improved, this method usually uses dual cameras or single cameras. The dual-camera dual-wavelength high-temperature measurement has high cost, complex structure, and poor integration. Although the single-camera dual-wavelength high-temperature measurement reduces the cost to a certain extent, similar to the dual-camera high-temperature measurement, it uses a beam splitter. The intensities of the two beams of light formed after a single wavelength enters the beam splitter need to be accurately calibrated, and using a beam splitter requires accurate matching of dual-wavelength images. However, the calibration of the beam splitter and the image matching error will affect the temperature detection accuracy. The dual-wavelength high-temperature measurement method based on a color camera avoids the problems of beam splitter calibration and image matching. However, this method is based on the assumption that pixels in adjacent regions correspond to the information of the same object point, which brings a large error to temperature measurement. Summary of the Invention

[0004] The present invention provides a method and system for in-situ monitoring of the molten pool temperature field based on multi-view light field (LF) imaging to solve the defects existing in the prior art and achieve high-precision temperature measurement without considering the pixel matching problem of different bands.

[0005] In a first aspect, the present invention provides a method for in-situ monitoring of the molten pool temperature field based on multi-view LF imaging, including: Obtaining multi-view LF signals of the molten pool; According to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, a Light Field Sub-aperture Bayer Model (LFSBM) is established, and the multi-view LF signal of the molten pool is input into the LFSBM to obtain a single-wavelength molten pool image containing R , G , B three channels; According to the two-wavelength temperature measurement theory, a blackbody furnace is used to calibrate the relative spectral responsivity of the LF camera; From R , G , B any two single-wavelength molten pool images and the relative spectral responsivity in the three channels, the temperature field of the molten pool is obtained based on the two-wavelength temperature measurement theory.

[0006] According to an in-situ monitoring method for the temperature field of a molten pool based on multi-view light field imaging provided by the present invention, according to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, an LFSBM is established, and the multi-view light field signal of the molten pool is input into the LFSBM to obtain a single-wavelength molten pool image containing R , G , B three channels, including: In the sensor pixel array under the Bayer filter, a coordinate system is established with the upper left corner of the pixel array as the origin, and the distribution law of the pixel RGB channels is obtained:

[0007] In the formula, ([[]] u , v ) is the pixel position index, \ is the remainder operation. When the sensor pixel array is distributed in the GRBG mode, and the starting point in the upper left corner is G and R , if the pixel row and column indices ( u , v ) = (odd, even), the pixel at this position is the pixel brightness value corresponding to the red filter, if ( u , v ) = (even, odd), it is the pixel brightness value corresponding to the blue filter, otherwise it is the pixel brightness value corresponding to the green filter; Determine that each microlens unit A in the LF camera covers a macro image pixel block B , and extract the A from the i st row to the m rd row, and the j th column to the n th column of the microlensA ij , and the corresponding macro image B ij , extract the pixels of the same wavelength from each macro image and splice them in sequence to obtain a single-wavelength molten pool image:

[0008] In the formula, I λ is λ the molten pool image at the wavelength of, m , n are the numbers of microlenses in the row and column, P i,j ( λ ) represents i , j the pixel value at, represents the splicing operation; Obtain the center coordinates of the microlenses by calibrating the white image with the LF camera tool, and correspond the center coordinates of the microlenses with the filter values in the filter distribution rule one by one, to obtain the distribution information of the center pixel values of the microlenses in four cases of two modes centered on R , B and centered on G .

[0009] According to an in-situ monitoring method for the molten pool temperature field based on multi-view light field imaging provided by the present invention, according to the dual-wavelength temperature measurement theory, a blackbody furnace is used to calibrate the relative spectral responsivity of the LF camera, including: Determine the dual-wavelength temperature measurement formula as:

[0010] In the formula, T represents the temperature, C 2 is a constant, λ is the wavelength, M1 and M2 are the gray values of two different wavelengths detected by the camera. Let k = Ln(r 2 / r 1 ) , r 1 and r 2 represent the relative spectral responsivities of the camera to the light beams of two different wavelengths; Adopt RB , RG , BG channel combinations for calibration analysis, and use a blackbody furnace at a working distance of LCalibrate the relative spectral responsivity of the LF camera. Place a neutral density filter between the LF camera and the blackbody furnace to reduce the light intensity. In the temperature range T 1 to T 2 Calibrate within the range. The blackbody furnace starts from T 1 temperature, and increases gradually by Δ T to T 2 temperature. Record the gray-scale ratio of the two wavelengths at different temperatures. According to the current temperature value T , calculate the parameter k , and perform a linear fit in combination with the temperature to determine the relative spectral responsivity of the LF camera.

[0011] Specifically, with temperature T as the vertical axis and the relative spectral responsivity k as the horizontal axis, perform a linear regression by the least squares method to obtain the calibration result of the relative spectral responsivity of the LF camera. Through linear regression, T and k establish a quadratic linear relationship:

[0012] Among them, S , U , E , W , F , Y , Q , H , D are constant coefficients, which are determined by T 1 - T 2 temperature and can be obtained by performing a linear regression by the least squares method on the blackbody furnace calibration result. k bg represents from B , G channel, k rb represents from R , B channel, k rg represents from R , G channel, and is the relative spectral responsivity derived from the two-wavelength temperature measurement formula.

[0013] In the second aspect, the present invention also provides an integrated application for in-situ monitoring of the molten pool temperature field in laser metal additive manufacturing based on multi-view light field imaging, including: a system and a coaxial system; Applications of the off-axis system include Laser Directed Energy Deposition (LDED) devices, and applications of the coaxial system include Laser Powder Bed Fusion (LPBF) devices.

[0014] According to an integrated application of an in-situ monitoring system for a molten pool temperature field based on multi-view light field imaging provided by the present invention, the LDED device includes an LF camera, a neutral density filter, and a laser processing system; The laser emitted by the laser processing system melts the object to be processed to form a molten pool. The light signal of the molten pool reaches the LF camera through the neutral density filter, and the LF camera transmits the molten pool image data formed by the light signal of the molten pool to the image processing system.

[0015] According to an integrated application of an in-situ monitoring system for a molten pool temperature field based on multi-view light field imaging provided by the present invention, the LPBF device includes an LF camera, a neutral density filter, a dichroic mirror, a field lens, a galvanometer scanner, and a laser processing system; The laser emitted by the laser processing system passes through the beam expander, the field lens, and the galvanometer scanner in sequence and then reaches the surface of the object to be processed to form a molten pool; The light signal of the molten pool passes through the field lens, the galvanometer scanner to the dichroic mirror, and through the neutral density filter to the LF camera. The LF camera transmits the molten pool image data formed by the light signal of the molten pool to the image processing system.

[0016] In a third aspect, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the in-situ monitoring method for a molten pool temperature field based on multi-view light field imaging as described in any one of the above.

[0017] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the in-situ monitoring method for a molten pool temperature field based on multi-view light field imaging as described in any one of the above.

[0018] In a fifth aspect, the present invention further provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the in-situ monitoring method for a molten pool temperature field based on multi-view light field imaging as described in any one of the above.

[0019] Compared with the prior art, the beneficial effects of the present invention include: 1) Greatly simplifies the monitoring process. Only single - time calibration of relative spectral responsivity is required, avoiding the image matching and spectral ratio calibration processes of traditional dual - wavelength temperature measurement systems, and overcoming the difficulty of reduced temperature measurement accuracy caused by matching and calibration errors. At the same time, compared with traditional color camera temperature measurement technology, its multi - perspective information improves the temperature measurement accuracy; 2) Has high temperature measurement accuracy and low error. Verified by a blackbody furnace, taking 30×30 pixels as an example, the temperature field error remains below 3%. Among them, the average error at the highest temperatures of 2973.15K, 3073.15K, and 3273.15K is 1.03%; 3) Compared with existing temperature measurement methods, the method proposed in the present invention has the advantages of a wide temperature range and low complexity compared with traditional infrared cameras and new dual - wavelength measurement methods; 4) The multi - camera temperature field in - situ monitoring technology using LF cameras has broad application prospects for in - situ monitoring of high - temperature metal molten pools during the additive manufacturing process. It can provide a relatively wide temperature range, the system is compact and easy to integrate. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following - described drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0021] Figure 1 is a schematic flow chart of the in - situ monitoring method of the molten pool temperature field based on multi - camera LF imaging provided by the present invention; Figure 2 is a schematic principle diagram of the application of the paraxial system, such as the multi - camera LF in - situ monitoring device of the temperature field in an LDED device; Figure 3 is a schematic principle diagram of the application of the coaxial system, such as the multi - camera LF in - situ monitoring device of the temperature field in an LPBF device; Figure 4 is a schematic diagram of the multi - camera imaging principle of the LF camera provided by the present invention; Figure 5 is a mapping relationship diagram of the Bayer matrix of the LF camera provided by the present invention; Figure 6 is a schematic diagram of the coordinates of the Bayer pixel block provided by the present invention; Figure 7 is a distribution diagram of four modes of the Bayer array provided by the present invention; Figure 8 is a pixel image extracted by the same filter provided by the present invention; Figure 9It is the relative spectral responsivity diagram of the LF camera calibrated by the blackbody furnace provided by the present invention; Figure 10 It is the monitoring result diagram of the molten pool temperature of TiAl4V alloy by using the LPBF coaxial monitoring system provided by the present invention; Figure 11 It is the structural schematic diagram of the electronic device provided by the present invention. Specific embodiments

[0022] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] In view of the problems existing in the prior art, the present invention proposes a new method for in-situ monitoring of the temperature field of the additive manufacturing molten pool by using the LF camera multi-view system.

[0024] Figure 1 It is the flow schematic diagram of the in-situ monitoring method of the molten pool temperature field based on multi-view light field imaging provided by the embodiment of the present invention, as Figure 1 shown, including: Step 100: Obtain the multi-view LF signals of the molten pool; Step 200: According to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, establish the LFSBM, input the multi-view LF signals of the molten pool into the LFSBM, and obtain a single-wavelength molten pool image including R , G , B three channels; Step 300: According to the dual-wavelength temperature measurement theory, calibrate the relative spectral responsivity of the LF camera by using a blackbody furnace; Step 400: From R , G , B any two single-wavelength molten pool images and the relative spectral responsivity in the three channels, obtain the molten pool temperature field based on the dual-wavelength temperature measurement theory.

[0025] Specifically, the monitoring system proposed by the embodiment of the present invention includes the application of a metal additive manufacturing processing system and a multi-view LF temperature monitoring system, including an off-axis and a coaxial system. The multi-view LF temperature monitoring system includes an LF camera, a neutral density filter, and an image processing system. The present invention adopts the LF camera multi-view system and obtains R , G , BPool images of three channels are used to calibrate the relative spectral responsivity of the LF camera. Starting from R , G , B Two-channel pool images are selected from the pool images of the three channels, and combined with the calibration parameters, the pool temperature field is deduced by the two-wavelength temperature measurement method. A blackbody furnace is used to calibrate the LF optical field temperature measurement system, and the monitoring temperature field error is kept below 3%. The present invention simplifies the monitoring process, improves the temperature measurement accuracy, has a wide temperature range and low complexity, can perform in-situ monitoring on high-temperature metal pools, and has broad application prospects in the field of additive manufacturing.

[0026] As Figure 2 and Figure 3 shown, the present invention uses an LF camera to construct an in-situ monitoring system. The paraxial system is applied to processes such as laser directed energy deposition, and the coaxial system is applied to processes such as laser powder bed fusion. The in-situ monitoring system includes an LF camera, a neutral density filter, a dichroic mirror for the application of the coaxial system such as the LPBF process, a field lens, a galvanometer, an image processing system, and a laser processing system. For the application of the paraxial system such as the LDED process, the pool optical signal directly reaches the LF camera through the neutral density filter, and the LF camera transmits the pool image data to the image processing system; for the application of the coaxial system such as the LPBF process, the pool optical signal passes through the field lens, the galvanometer to the dichroic mirror, and then through the neutral density filter to the LF camera. The LF camera transmits the multi-view LF signal of the pool to the image processing system. The neutral density filter is used to adjust the light intensity to prevent the camera from overexposing.

[0027] An in-situ monitoring method for the pool temperature field during metal additive manufacturing based on multi-view LF imaging of the above system includes the following steps: Step 1: Establishment of LFSBM. By analyzing the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, LFSBM is established to obtain the pool images of single-wavelength channels including R , G , B three channels.

[0028] Preferably, in step 1, LFSBM is specifically to obtain the center coordinates of the microlens by calibrating the white image, match with the color filter array, determine the pixel value distribution under different Bayer patterns, and thus obtain the single-wavelength images of the pool of different channels.

[0029] The two-wavelength temperature measurement formula is as follows: (1) In the formula, T represents temperature, C 2 is a constant, λ is the wavelength, MLet k = Ln ( r 2 / r 1 ), r 1 and r 2 represent the relative spectral responsivities of the camera to light beams of different wavelengths.

[0030] Among them, the principle of multi-view imaging is as follows: As Figure 4 shown, the LF camera captures the light rays from object points at different angles through a microlens array and images them to different positions on the sensor, as Figure 4 shown by the red frame in P , where object points at different angles pass through different microlens units and are imaged to different pixel positions under the microlens units on the sensor, thus retaining the

[0031] angle information of the target points. R , G , B Generally, a color camera is equipped with R , G , B color filters in front of the sensor to capture the original Bayer format image, and a common color image is obtained through a post-interpolation algorithm. To obtain images of different wavelengths, the LF camera used in the present invention has built-in GRBG color filters. Taking the image in the format of " Figure 5 " as an example, such as A LFSBM shown in B , each microlens unit A covers a pixel block called a macro image i . Select the microlenses from the m th row to the j th row and from the n th column to the A ij th column in i for analysis. j represents the microlens in the B ij th row and the th column, corresponding to the macro image GRBG it covers. Each pixel in the macro image has a different and unknown wavelength distribution. (2) In the formula, ( u , v ) is the pixel position index, \ is the modulo operation. If the sensor pixel distribution is GRBG mode, and the starting point is at the upper left corner as G and R , when the pixel row and column indices ( u , v ) = (odd, even), the pixel at this position is the pixel brightness value corresponding to the red filter. When ( u , v ) = (even, odd), it is the pixel brightness value corresponding to the blue filter, otherwise it is the pixel brightness value corresponding to the green filter; Since the temperature of the molten pool is the same under different viewing angles, to obtain the image of the scene under the same filter, based on each macro image being the information of the same object point from different viewing angles, the pixels of the same wavelength can be taken from each macro image and stitched in order to obtain images of different wavelengths. Therefore, it can be defined as: (3) In the formula, I λ is the image at wavelength λ, m , n are the numbers of microlenses in the row and column, P i,j ([[]] λ ) represents i , j is the pixel value at represents the stitching operation.

[0032] The center coordinates of the microlenses are obtained by calibrating the white image with the LF camera tool, and the center coordinates of the microlenses are corresponding to the filter values in the filter distribution rule one by one, obtaining the distribution information of the center pixel values of the microlenses in a total of four cases of two modes centered on R , B and centered on G .

[0033] As Figure 7 shown, the four modes are the R distribution centered on RGGB (mode 1), the B distribution centered on BGGR (mode 2), the G distribution centered on GRBG (mode 3), and the G distribution centered on GBRGDistribution (Pattern 4). In the present invention, 25 viewing angles are marked with numbers in each distribution pattern. To distinguish different distribution patterns, the numbers of each pattern have different colors, and Patterns 1-4 are represented by white, black, yellow, and gray respectively. Figure 8 Single-wavelength images extracted from the same filter. Among them, R 1 Is represented as the splicing combination of the pixels indexed 13, 7, 12, and 8 in the macro-images at positions ( i , j ), ( i , j + 1), ( i + 1, j ), and ( i + 1, j + 1). R 2 - R 4 , B 1 - B 4 , G 1 - G 4 Similarly. Finally, the single-wavelength molten pool images of R , G , B can be obtained.

[0034] Step 2: Relative spectral responsivity calibration of the LF camera. According to the two-wavelength temperature measurement theory, a blackbody furnace is used to calibrate the relative spectral responsivity of the LF camera.

[0035] Preferably, in Step 2, a specific temperature range of the blackbody furnace is selected for calibration and verification. By recording the gray-scale ratio of the two wavelengths at different temperatures, the parameter k in Equation (1) is obtained, and the relative spectral responsivity of the LF camera is determined by linear fitting in combination with the temperature.

[0036] According to Equation (1), based on the single-wavelength sub-aperture images outlined in Step 1, calibration analysis is performed using the channel combinations of ( R 1 , B 1 ), ( R 1 , G 1 ), ( B 1 , G 1 ). Using the blackbody furnace at a working distance of LCalibrate the relative spectral responsivity of the LF camera. Place a neutral density filter between the LF camera and the blackbody furnace to reduce the light intensity. Select T 1 to T 2 for calibration. The blackbody furnace starts from T 1 temperature and gradually increases by Δ P to T 2 temperature. Through this process, record the gray-scale ratio of the two wavelengths at different temperatures. According to the current temperature value T , calculate the parameter k in formula (1), and perform linear fitting in combination with the temperature to determine the relative spectral responsivity of the LF camera.

[0037] Furthermore, the relative spectral responsivity is calibrated as follows: Taking the calibration result in the temperature range of the blackbody furnace T 1 - T 2 as an example, with temperature T as the vertical axis and the relative spectral responsivity k as the horizontal axis, perform linear regression by the least squares method to obtain the calibration result of the relative spectral responsivity of the LF camera. Among them, k bg represents the ratio derived from formula (1) from B , G channel, k rb represents the ratio derived from formula (1) from R , B channel, k rg represents the ratio derived from formula (1) from R , G channel. Through linear regression, a quadratic linear relationship is established between T and k : (4) Among them, in the formula S , U , E , W , F , Y , Q , H , D are constant coefficients, which are determined by the T 1 - T 2 temperature and can be obtained by performing linear regression by the least squares method on the blackbody furnace calibration result. When T I= 1273.15 K, T 2 = 2893.15 K, the calibration results of the relative spectral responsivity of the LF camera by the blackbody furnace are as Figure 9 shown. The root mean square errors of the temperature fitting are 76.34 K, 62.24 K, and 78.66 K respectively.

[0038] Step 3, obtain the molten pool temperature field. Based on R , G , B the dual-wavelength images and the relative spectral responsivity at the centers of two channels among the three channels, obtain the molten pool temperature field based on the dual-wavelength temperature measurement theory.

[0039] Specifically, taking the LPBF coaxial monitoring system as an example, the present invention monitors the molten pool of Ti6Al4V alloy. Through the established LFSBM, obtain R , G , B the molten pool images of the three channels, combined with the LF camera after blackbody furnace calibration, use the dual-wavelength temperature measurement method to deduce the molten pool temperature field, and its temperature field distribution is as Figure 10 shown.

[0040] Figure 11 Illustrates a schematic diagram of the physical structure of an electronic device, as Figure 11 shown. The electronic device may include: a processor 1110, a communication interface 1120, a memory 1130, and a communication bus 1140. Among them, the processor 1110, the communication interface 1120, and the memory 1130 complete mutual communication through the communication bus 1140. The processor 1110 can call the logical instructions in the memory 1130 to execute the in-situ monitoring method for the molten pool temperature field based on multi-view LF imaging. The method includes: obtaining the multi-view LF signals of the molten pool; according to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, establish the LFSBM, input the multi-view LF signals of the molten pool into the LFSBM, and obtain the sub-aperture images and the single-wavelength molten pool images including R , G , B the three channels; according to the dual-wavelength temperature measurement theory, calibrate the relative spectral responsivity of the LF camera using the blackbody furnace, and from R , G , B any two single-wavelength molten pool images and the relative spectral responsivity among the three channels, obtain the molten pool temperature field based on the dual-wavelength temperature measurement theory.

[0041] In addition, when the logical instructions in the above-mentioned memory 1130 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0042] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the in-situ monitoring method of the molten pool temperature field based on multi-view LF imaging provided by the above-mentioned various methods. The method includes: acquiring multi-view LF signals of the molten pool; according to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, establishing an LFSBM, inputting the multi-view LF signals of the molten pool into the LFSBM, and obtaining sub-aperture images including R , G , B three channels and a single-wavelength molten pool image; according to the dual-wavelength temperature measurement theory, using a blackbody furnace to calibrate the relative spectral responsivity of the LF camera, and from R , G , B any two single-wavelength molten pool images and the relative spectral responsivity in the three channels, obtaining the molten pool temperature field based on the dual-wavelength temperature measurement theory.

[0043] On another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it realizes the in-situ monitoring method of the molten pool temperature field based on multi-view LF imaging provided by the above-mentioned various methods. The method includes: acquiring multi-view LF signals of the molten pool; according to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, establishing an LFSBM, inputting the multi-view LF signals of the molten pool into the LFSBM, and obtaining sub-aperture images including R , G , B three channels and a single-wavelength molten pool image; according to the dual-wavelength temperature measurement theory, using a blackbody furnace to calibrate the relative spectral responsivity of the LF camera, and from R ,G , B The single-wavelength molten pool images and relative spectral responsivities of any two of the three channels, and the molten pool temperature field is obtained based on the dual-wavelength temperature measurement theory.

[0044] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0045] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0046] 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 described in the foregoing embodiments or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for in-situ monitoring of molten pool temperature field based on multi-view field imaging, characterized in that: include: Acquire multi-view light field LF signals of the molten pool; According to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, a light field sub-aperture Bayer model LFSBM is established, and the multi-view LF signal of the molten pool is input into the LFSBM to obtain R , G , B Sub-aperture images of three channels and single-wavelength melt pool images; According to the dual-wavelength temperature measurement theory, a blackbody furnace is used to calibrate the relative spectral responsivity of the LF camera; Depend on R , G , B Any two single-wavelength molten pool images and relative spectral responsivities in the three channels are used to obtain the molten pool temperature field based on the dual-wavelength temperature measurement theory.

2. The in-situ monitoring method of the molten pool temperature field based on multi-view field imaging according to claim 1 is characterized in that: According to the mapping relationship between the microlens unit and the Bayer pixel block of the LF camera, an LFSBM is established, and the multi-view LF signal of the molten pool is input into the LFSBM to obtain R , G , B Sub-aperture images of three channels and single-wavelength melt pool images, including: In the sensor pixel array under the Bayer filter, the pixel coordinate system is established with the upper left corner of the pixel array as the origin, and the pixel RGB channel distribution law is obtained: In the formula, ( u , v ) is the pixel position index, \ is the remainder operation, if the sensor pixel array is distributed as GRBG Mode, and the upper left corner starts at G and R , when the pixel row and column index ( u , v ) = (odd number, even number), the pixel at this position is the pixel brightness value corresponding to the red light filter. u , v ) = (even number, odd number), it is the pixel brightness value corresponding to the blue light filter, otherwise it is the pixel brightness value corresponding to the green light filter; Determine each microlens unit in the LF camera A Overlay a macro image pixel block B ,extract A Middle i Go to m Row, No. j Column to n Microlenses A ij , and the corresponding macro image B ij , extract pixels of the same wavelength from each macro image and sequentially stitch them together to obtain images of different wavelengths: In the formula, I λ is the image at wavelength λ, m , n is the number of microlenses in rows and columns, P i,j ( λ )express i , j The pixel value at Represents a splicing operation; The white image is calibrated by the LF camera tool to obtain the center coordinates of the microlens, and the center coordinates of the microlens are matched one by one with the filter values ​​in the filter distribution law to obtain the following: R , B Centered on G The pixel value distribution information of the microlens center in four cases with two modes in the center.

3. The in-situ monitoring method of the molten pool temperature field based on multi-view field imaging according to claim 2 is characterized in that: According to the dual-wavelength temperature measurement theory, a blackbody furnace is used to calibrate the relative spectral responsivity of the LF camera, including: The dual-wavelength temperature measurement formula is determined as follows: In the formula, T Indicates temperature, C 2 is a constant, λ is the wavelength, M 1 and M 2 is the gray value of two different wavelengths detected by the camera, let k = Ln(r 2 / r 1 ) , r 1 and r 2 represents the relative spectral responsivity of the camera to two different wavelength beams; use RB , RG , BG The channel combination is calibrated and analyzed using a blackbody furnace at a working distance of L The relative spectral responsivity of the LF camera is calibrated above, and a neutral density filter is placed between the LF camera and the black body furnace to reduce the light intensity; In temperature range T 1 to T 2 range, the blackbody furnace is calibrated from T 1 temperature, every Δ T Gradually increase to T 2 Temperature, record the grayscale ratio of dual wavelengths at different temperatures, according to the current temperature value T , calculation parameters k , and a linear fit is performed in combination with the temperature to determine the relative spectral responsivity of the LF camera.

4. The method for in-situ monitoring of molten pool temperature field based on multi-view field imaging according to claim 3 is characterized in that: According to the dual-wavelength temperature measurement theory, a blackbody furnace is used to calibrate the relative spectral responsivity of the LF camera, which also includes: By temperature T The vertical axis is the relative spectral responsivity k As the horizontal axis, the linear regression is performed by the least square method to obtain the relative spectral response calibration result of the LF camera. T and k A quadratic linear relationship is established between them: in, S , U , E , W , F , Y , Q , H , D is a constant coefficient, given by T 1- T 2 temperature, which can be obtained by performing least squares linear regression on the blackbody furnace calibration results. k bg Indicates from B , G aisle, k rb Indicates from R , B aisle, k rg Indicates from R , G The channels use a ratio derived from the dual wavelength thermometry formula.

5. A molten pool temperature field in-situ monitoring system based on multi-view field imaging, used for executing the image processing system in the molten pool temperature field in-situ monitoring method based on multi-view field imaging as described in any one of claims 1 to 4, characterized in that: include: Rangefinder and coaxial systems; The applications of paraxial systems include laser directed energy deposition (LDED) equipment, and the applications of coaxial systems include laser powder bed fusion (LPBF) equipment.

6. The in-situ monitoring system for molten pool temperature field based on multi-view field imaging according to claim 5 is characterized in that: LDED equipment includes LF cameras, neutral density filters, and laser processing systems; The laser emitted by the laser processing system melts the object to be processed to form a molten pool, and the molten pool light signal reaches the LF camera through the neutral density filter. The LF camera transmits the molten pool image data formed by the molten pool light signal to the image processing system.

7. The in-situ monitoring system for molten pool temperature field based on multi-view field imaging according to claim 5 is characterized in that: LPBF equipment includes LF camera, neutral density filter, dichroic mirror, field lens, galvanometer and laser processing system; The laser emitted by the laser processing system passes through the beam expander, the field lens, and the galvanometer in sequence and reaches the surface of the object to be processed to be melted to form a molten pool; The molten pool light signal passes through the field lens and the galvanometer to the dichroic mirror, and then passes through the neutral density filter to the LF camera. The LF camera transmits the molten pool image data formed by the molten pool light signal to the image processing system.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the in-situ monitoring method of the molten pool temperature field based on multi-view field imaging as described in any one of claims 1 to 4 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the in-situ monitoring method of the molten pool temperature field based on multi-view field imaging as described in any one of claims 1 to 4 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the in-situ monitoring method of the molten pool temperature field based on multi-view field imaging as described in any one of claims 1 to 4 is implemented.

Images