Silicon light FA alignment equipment and alignment method thereof

Abstract

The invention discloses silicon light FA alignment equipment and an alignment method thereof. The silicon light FA alignment equipment comprises a product carrying table mechanism, and a UV curing mechanism and an alignment structure are arranged on the periphery of the product carrying table mechanism; the product carrying table mechanism comprises a carrying table arranged on the carrying table leveling mechanism; the alignment structure comprises a clamping driving mechanism arranged on the alignment posture adjusting platform deck, and a clamping mechanism is arranged on the clamping driving mechanism; the UV curing mechanism comprises a UV curing assembly and an image detection auxiliary mechanism which are arranged on the displacement driving mechanism; the clamping mechanism obtains an FA device, the FA device can move relative to a product assembly on the carrying table through the clamping driving mechanism and the clamping mechanism, and position verification and curing assembly are achieved through a UV curing assembly in the UV curing mechanism and the image detection auxiliary mechanism. The invention discloses a silicon light FA alignment device and an alignment method thereof, which realize real-time alignment of a silicon light FA product through mutual cooperation of alignment, image assistance, UV curing and a product carrying table mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of welding in mechanical automation, and in particular to a silicon photonic FA alignment device and an alignment method thereof. Background Art

[0002] FA alignment refers to the process of connecting the fiber array (FA) to other optical components or devices. At present, the output of silicon photonics FA products is gradually increasing, and the efficiency of manual alignment and mounting of FA cannot meet the needs of production capacity. There is an urgent need to realize the automation of FA alignment and mounting. In view of the difficulty of manual alignment and value finding on the platform built by customers themselves, unstable value finding, time-consuming and labor-intensive value finding, and cumbersome statistical data, we have designed a completely automatic alignment and value finding and automatic mounting UV curing based on the relevant characteristics of the product FA, so as to reduce manual participation and realize fast, accurate and stable alignment and value finding, thereby improving the alignment and mounting efficiency of the product, product yield, saving manpower and reducing production costs.

[0003] Therefore, the following problems still exist in the prior art: the efficiency of alignment and placement is low; the alignment and placement process is complicated and highly dependent on personnel; the low degree of automation makes it difficult to meet large-scale production needs; and data statistics are complicated and difficult to quantify. Summary of the invention

[0004] The present invention overcomes the shortcomings of the prior art and provides a silicon photonics FA alignment device and an alignment method thereof. The real-time alignment of silicon photonics FA products is achieved through the cooperation of alignment, image assistance, UV curing, and a product stage mechanism.

[0005] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a silicon photonic FA alignment device, comprising a product stage mechanism for receiving and positioning product components, and a UV curing mechanism and an alignment structure are arranged on the periphery of the product stage mechanism; The product stage mechanism includes a stage leveling mechanism and a stage arranged on the stage leveling mechanism, and a product component is positioned on the stage; The alignment structure includes an alignment posture adjustment platform, a clamping drive mechanism is arranged on the alignment posture adjustment platform, and a clamping mechanism is arranged on the clamping drive mechanism; The UV curing mechanism includes a displacement driving mechanism, and a UV curing component and an image detection auxiliary mechanism arranged on the displacement driving mechanism; The clamping mechanism obtains the FA device, and adjusts the position of the FA device by the clamping drive mechanism and the displacement of the product component on the carrier to achieve alignment between the FA device and the product component, and realizes position verification and curing assembly through the UV curing component in the UV curing mechanism and the image detection auxiliary mechanism.

[0006] In a preferred embodiment of the present invention, the displacement drive mechanism includes an XY displacement mechanism and a longitudinal adjustment mechanism arranged on the XY displacement mechanism; the longitudinal adjustment mechanism is provided with a pair of relatively arranged angle adjustment brackets, and the angle adjustment brackets are provided with a pair of relatively arranged UV curing mechanisms.

[0007] In a preferred embodiment of the present invention, an image detection auxiliary mechanism is provided on the displacement driving mechanism, and the detection sampling port of the image detection auxiliary mechanism corresponds to the carrier in the product carrier mechanism.

[0008] In a preferred embodiment of the present invention, the clamping mechanism includes an adsorption head and a pair of clamping claws spaced apart on both sides of the adsorption head. The clamping claws are connected to the positioning posture adjustment platform through the clamping drive mechanism, and the clamping drive mechanism drives the clamping claws on both sides of the adsorption head to open and close the clamp relative to the adsorption head.

[0009] In a preferred embodiment of the present invention, a method for aligning a silicon photonic FA alignment device is implemented using a silicon photonic FA alignment device, and includes the following steps: The initial position between the FA device acquired by the clamping mechanism in the alignment structure and the product component carried by the product stage mechanism is acquired by an image detection auxiliary mechanism provided on the displacement driving mechanism in the UV curing mechanism; Based on the position image between the FA device and the product component obtained by the image detection auxiliary mechanism, the relative position relationship is obtained through the image processing algorithm, and the compensation parameters are obtained by the alignment algorithm based on the relative position relationship. The stage leveling mechanism in the product stage mechanism is driven according to the supplementary parameters to adjust the posture of the product component positioned on the stage, and the stage adjusts the posture of the FA device through the alignment posture adjustment structure to achieve alignment between the product component and the FA device, so that the product component and the FA device meet the pre-set alignment parameters.

[0010] In a preferred embodiment of the present invention, after dispensing, the aligned product components and FA devices are UV-cured by a UV curing component of a UV curing mechanism.

[0011] In a preferred embodiment of the present invention, the image processing algorithm comprises the following steps: The visual edge capture algorithm is used to process and analyze the images of chips and FA devices on the product components in the position image, and the object edge information of the chips and FA devices in the image is extracted. The object edge information includes position information and angle information.

[0012] In a preferred embodiment of the present invention, the visual edge grabbing algorithm comprises the following steps: Step a1, applying Gaussian filtering to smooth the position image and preliminarily remove noise from the position image; Step a2, finding the gradient of the position image; Step a3, applying a non-maximum suppression algorithm to the position image processed in step a2 based on the gradient of the position image to filter out non-edge pixels, making the blurred boundary clear; and retaining the maximum value of the gradient intensity on the pixel point to further filter out noise; Step a4, applying a double threshold method to determine potential boundaries; In step a5, the boundary is tracked using the turbulence post algorithm. When the weak boundary whose pixel position is connected to the strong boundary is considered to be the boundary, the other weak boundaries are deleted.

[0013] In a preferred embodiment of the present invention, in step a2, finding the gradient of the position image is: calculating the gradient magnitude and direction according to the Sobel operator, finding the gradient of the image, and then calculating the gradient magnitude and direction; And / or, in step a3, the non-maximum suppression algorithm is: approximate the gradient direction to one of the proposed values, then the gradient direction represents the up, down, left, right and 45-degree directions; compare the gradient strengths of the corresponding pixel points and the pixel points in the positive and negative directions of the gradient, and retain the pixel point when the gradient strength is the largest, otherwise suppress it; And / or, in step a4, the double threshold method is used to determine the potential boundary: an upper threshold and a lower threshold are set; when the pixel in the image is greater than the upper threshold, it is considered to be a boundary, which is called a strong boundary; when the pixel in the image is less than the lower threshold, it is considered not to be a boundary; when the pixel in the image is between the upper threshold and the lower threshold, it is considered to be a candidate, which is called a weak boundary; The alignment algorithm includes generalized polynomial fitting using the Bisquare method, and the best position is extracted after polynomial fitting of discrete data.

[0014] In a preferred embodiment of the present invention, the alignment algorithm includes the following steps: Angle adjustment between the chip and the FA device on the product component: Obtain the edge angle x of the chip and the edge angle y of the FA device through the visual edge capture algorithm in the second embodiment, and then calculate the number of rotation axis pulses z through the formula z=(xy)×m; wherein m is the coefficient of the angle to pulse number, and after the rotation axis rotates z pulses, the chip and the edge of the FA device can remain parallel; Through the alignment algorithm, the coordinate axis position corresponding to the peak of the curve during the alignment process is analyzed, and the stepper axis is controlled by the motion control card to align the optical paths of the chip and FA device to achieve the maximum optical power value.

[0015] Specifically, the polynomial function of the data fitting is: ; where f is the output sequence of the best polynomial fit, X is the input sequence, a is the polynomial coefficient, and m is the polynomial order.

[0016] Specifically, the Bisquare method is used to obtain the value of a for the best fitting observation point (X, Y), including the following steps: Use the least squares method to perform polynomial fitting to obtain the initial estimate of the polynomial coefficients: a 0 k , a 1 k , a 2 k , a 3 k …, a m k , k=0; update a 0 k , a 1 k , a 2 k , a 3 k …, a m k The weight W is used; the least squares method and the new weights are used to perform polynomial fitting to obtain a new set of polynomial coefficients: a 0 k+1 , a 1 k+1 , a 2 k+1 , a 3 k+1 …, a m k+1 ; when min(|△a 0 / a 0 …△a m / a m |)<Tolerance is true, output a 0 k+1 , a 1 k+1 , a 2 k+1 , a 3 k+1 …, a m k+1 , when min(|△a 0 / a 0 …△a m / a m |) < tolerance is false, then k 新 =k+1, jump to update a again 0 k , a 1 k , a 2 k , a 3 k …, a m kThe weight W is used, and the least squares method and the new weights are used to perform polynomial fitting to obtain a new set of polynomial coefficients.

[0017] The present invention solves the defects existing in the background technology and has the following beneficial effects: The present invention discloses a silicon photonics FA alignment device and an alignment method thereof, which realizes the alignment of silicon photonics FA products through the cooperation of a UV curing mechanism, an alignment mechanism, a product stage mechanism, and an image detection auxiliary mechanism, and is equipped with an alignment algorithm and image processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0019] Figure 1 It is an overall axial view of a silicon photonic FA alignment device according to a preferred embodiment of the present invention; Figure 2 It is a schematic diagram of the axial structure of a UV curing mechanism of a silicon photo FA alignment device according to a preferred embodiment of the present invention; Figure 3 It is a schematic diagram of the axial structure of an alignment mechanism of a silicon photonic FA alignment device according to a preferred embodiment of the present invention; Figure 4 The preferred embodiment of the present invention Figure 3 The enlarged structural diagram of the middle A part; Figure 5 This is a schematic diagram of the axial structure of a product stage mechanism of a silicon photonic FA alignment device according to a preferred embodiment of the present invention; Figure 6 A schematic diagram of the working process of a silicon photonic FA alignment device according to a preferred embodiment of the present invention; Figure 7 A schematic diagram of a workflow of using a non-maximum suppression operation in a silicon photonic FA alignment device in a preferred embodiment of the present invention to retain the maximum value of the gradient intensity on the pixel point and filter out other values; Figure 8 A schematic diagram of a work flow of using a dual threshold method to determine a possible (potential) boundary in a silicon photonic FA alignment device according to a preferred embodiment of the present invention; Fig. 9 A schematic diagram of a process of obtaining the a value of the best fitting observation point (X, Y) using the Bisquare method for a silicon photonic FA alignment device according to a preferred embodiment of the present invention.

[0020] Fig.10 Schematic diagram of using binary method to obtain boundary line of a silicon photonic FA alignment device according to a preferred embodiment of the present invention Figure 1 .

[0021] Fig.11Schematic diagram of using binary method to obtain boundary line of a silicon photonic FA alignment device according to a preferred embodiment of the present invention Figure 2 .

[0022] In the figure: 1-operating table, 2-UV curing mechanism, 21-displacement driving mechanism, 22-UV curing component, 23-image detection auxiliary mechanism, 3-alignment mechanism, 31-position posture adjustment platform, 32-clamping driving mechanism, 33-clamping mechanism, 331-clamping claw, 332-adsorption head, 4-product platform mechanism, 41-platform leveling mechanism, 42-platform, 43-product component, 5-FA device. DETAILED DESCRIPTION

[0023] The present invention will now be further described in detail with reference to the accompanying drawings and embodiments. These drawings are simplified schematic diagrams that only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention. Embodiment 1

[0024] like Figure 1-Figure 5 As shown, a silicon photonic FA alignment device includes a product carrier mechanism 4 for receiving and positioning a product component 43, and a UV curing mechanism 2 and an alignment structure 3 are arranged on the periphery of the product carrier mechanism 4.

[0025] Specifically, the product carrier mechanism 4 includes a carrier leveling mechanism 41 and a carrier 42 disposed on the carrier leveling mechanism 41 , and a product component 43 is positioned on the carrier 42 .

[0026] Specifically, the alignment structure 3 includes an alignment posture adjustment platform 31 , a clamping drive mechanism 32 is disposed on the alignment posture adjustment platform 31 , and a clamping mechanism 33 is disposed on the clamping drive mechanism 32 .

[0027] Specifically, the UV curing mechanism 2 includes a displacement driving mechanism 21, and a UV curing component 22 and an image detection auxiliary mechanism 23 arranged on the displacement driving mechanism 21. Further, the displacement driving mechanism 21 includes an XY displacement mechanism, and a longitudinal adjustment mechanism arranged on the XY displacement mechanism; a pair of angle adjustment brackets arranged oppositely are arranged on the longitudinal adjustment mechanism, and a pair of UV curing mechanisms 2 arranged oppositely are arranged on the angle adjustment bracket. The displacement driving mechanism 21 is provided with an image detection auxiliary mechanism 23, and the detection sampling port of the image detection auxiliary mechanism 23 corresponds to the stage 42 in the product stage mechanism 4.

[0028] Specifically, the clamping mechanism 33 obtains the FA device 5, and can be displaced relative to the product component 43 on the carrier 42 through the clamping drive mechanism 32 and the clamping mechanism 33, and realizes position verification and curing assembly through the UV curing component 22 and the image detection auxiliary mechanism 23 in the UV curing mechanism 2. Further, the clamping mechanism 33 includes an adsorption head 332, and a pair of clamping claws 331 arranged at intervals on both sides of the adsorption head 332, the clamping claws 331 are connected to the alignment posture adjustment carrier 31 through the clamping drive mechanism 32, and the clamping drive mechanism 32 drives the clamping claws 331 on both sides of the adsorption head 332 to open and close relative to the adsorption head 332. Embodiment 2

[0029] like Figure 1-Figure 9 As shown, an alignment method of a silicon photonics FA alignment device is implemented using the silicon photonics FA alignment device of embodiment 1, and includes the following steps: The initial position between the FA device 5 acquired by the clamping mechanism 33 in the alignment structure 3 and the product component 43 carried by the product stage mechanism 4 is acquired by the image detection auxiliary mechanism 23 provided on the displacement driving mechanism 21 in the UV curing mechanism 2; Based on the position image between the FA device 5 and the product component 43 obtained by the image detection auxiliary mechanism 23, the relative position relationship is obtained through the image processing algorithm, and the compensation parameters are obtained by the alignment algorithm based on the relative position relationship. The stage leveling mechanism 41 in the product stage mechanism 4 is driven according to the supplementary parameters to adjust the posture of the product component 43 positioned on the stage 42, and the posture of the FA device 5 is adjusted by the alignment posture adjustment stage 31 of the alignment structure 3 to achieve alignment between the product component 43 and the FA device 5, so that the product component 43 and the FA device 5 meet the pre-set alignment parameters.

[0030] Furthermore, glue is dispensed by the glue dispensing mechanism, and the aligned product component 43 and the FA device 5 are photocured by the UV curing component 22 of the UV curing mechanism 2 .

[0031] Specifically, the image processing algorithm includes the following steps: The image of the chip and FA device 5 on the product component 43 in the position image is processed and analyzed using a visual edge capture algorithm to extract the information of the object edge of the chip and FA device 5 in the image. The object edge information includes position information and angle information. The position of the object edge in the image is obtained by identifying the change in the image gray value. The contour edge of the object is fitted by continuously identifying the positions of multiple consecutive edge points and then performing curve fitting.

[0032] Among them, the visual edge capture algorithm includes the following steps: Step a1, applying Gaussian filtering to smooth the position image and preliminarily remove noise from the position image.

[0033] Since edge detection is susceptible to noisy images, denoising is usually required before edge detection. Use Gaussian filtering to remove noise. For example, the kernel size of the Gaussian filter is 5x5, and the formula includes: .

[0034] Step a2, finding the gradient of the position image.

[0035] Specifically, in step a2, the gradient of the position image is found by calculating the gradient magnitude and direction according to the Sobel operator, finding the gradient of the image, and then calculating the gradient magnitude and direction; The formula includes: ; ; ; ; Where S is the gradient magnitude and θ is the direction.

[0036] Step a3, based on the gradient of the position image, the non-maximum suppression algorithm is applied to the position image processed in step a2 to filter out non-edge pixels, so as to make the blurred boundary clear; and the maximum value of the gradient intensity on the pixel point is retained to further filter the noise.

[0037] Specifically, in step a3, the non-maximum suppression algorithm is: approximate the gradient direction to one of the proposed values, the proposed values ​​include 0, 45, 90, 135, 180, 225, 270, and 315, and the proposed value is the direction angle value; then the gradient direction represents the up, down, left, right and 45 degree directions; compare the gradient strength of the pixel points in the positive and negative directions of the gradient of the corresponding pixel points and the pixel points, and retain the pixel point when the gradient strength is the largest, otherwise suppress (delete, that is, set to 0).

[0038] In step a4, a double threshold method is applied to determine potential boundaries.

[0039] Specifically, in step a4, the double threshold method is applied to determine the potential boundary: set an upper threshold and a lower threshold; when the pixel in the image is greater than the upper threshold, it is considered to be a boundary, which is called a strong boundary; when the pixel in the image is less than the lower threshold, it is considered not to be a boundary; when the pixel in the image is between the upper threshold and the lower threshold, it is considered to be a candidate, which is called a weak boundary.

[0040] In step a5, the boundary is tracked using the turbulence post-processing technique. When the weak boundary whose pixel position is connected to the strong boundary is considered as the boundary, the other weak boundaries are deleted.

[0041] Specifically, the alignment algorithm includes the following steps: using the generalized polynomial fitting of the Bisquare method, performing polynomial fitting on the discrete data, and then extracting the best position.

[0042] Specifically, the polynomial function of the data fitting is: ; Where f is the output sequence of the best polynomial fit, X is the input sequence, a is the polynomial coefficient, and m is the polynomial order; Specifically, the Bisquare method is used to obtain the value of a for the best fitting observation point (X, Y), including the following steps: Use the least squares method to perform polynomial fitting to obtain the initial estimate of the polynomial coefficients: a 0 k , a 1 k , a 2 k , a 3 k …, a m k , k=0; update a 0 k , a 1 k , a 2 k , a 3 k …, a m k The weight W is used; the least squares method and the new weights are used to perform polynomial fitting to obtain a new set of polynomial coefficients: a 0 k+1 , a 1 k+1 , a 2 k+1 , a 3 k+1 …, a m k+1 ; when min|△a 0 / a 0 …△a m / a m |<When tolerance is true, output a 0 k+1 , a 1 k+1 , a 2 k+1 , a 3 k+1 …, a m k+1 , when min|△a 0 / a 0 …△a m / a m |<Tolerance is false, then k 新 =k+1, jump to update a again 0 k , a1 k , a 2 k , a 3 k …, a m k The weight W is used, and the least squares method and the new weights are used to perform polynomial fitting to obtain a new set of polynomial coefficients. Embodiment 3

[0043] On the basis of the second embodiment, the angle between the chip and the FA device 5 on the product component 43 is adjusted: the edge angle x of the chip and the edge angle y of the FA device 5 are obtained by the visual edge capture algorithm (image edge recognition method) in the second embodiment, and then the number of rotation axis pulses z is calculated by the formula z=(xy)×m. Where m is the coefficient of the angle to pulse number. After the rotation axis rotates by z pulses, the chip and the FA edge can remain parallel.

[0044] The alignment algorithm is used to analyze the coordinate axis position corresponding to the peak of the curve during the alignment process, and the stepping axis is controlled by the motion control card to align the optical paths of the chip and the FA device 5 to achieve the maximum optical power value.

[0045] Working principle: like Figure 1-Figure 9 As shown, the present invention discloses a silicon photonics FA alignment device and an alignment method thereof, which realizes the alignment of silicon photonics FA products through the cooperation of a UV curing mechanism 2, an alignment mechanism 3, a product stage mechanism 4, and an image detection auxiliary mechanism 23, and is equipped with an alignment algorithm and image processing. The alignment method of the silicon photonic FA alignment device of the present invention is implemented by using the silicon photonic FA alignment device: the initial position between the FA device 5 obtained by the clamping mechanism 33 in the alignment structure 3 and the product component 43 carried on the product stage mechanism 4 is obtained by the image detection auxiliary mechanism 23 set on the displacement driving mechanism 21 in the UV curing mechanism 2; the relative position relationship is obtained by the image processing algorithm according to the position image between the FA device 5 and the product component 43 obtained by the image detection auxiliary mechanism 23, and the compensation parameters are obtained by the alignment algorithm according to the relative position relationship, and the stage leveling mechanism 41 in the product stage mechanism 4 is driven to adjust the posture of the product component 43 carried and positioned on the stage 42 according to the supplementary parameters, and the posture of the FA device 5 is adjusted by the alignment posture adjustment stage 31 of the alignment structure 3, so as to realize the alignment (or coupling) of the product component 43 and the FA device 5, so that the product component 43 and the FA device 5 meet the preset alignment parameters. In addition, the glue is dispensed by the glue dispensing mechanism, and the aligned product component 43 and the FA device 5 are light-cured by the UV curing component 22 of the UV curing mechanism 2.

[0046] The present invention uses visual edge capture technology to process and analyze the images of the chip and FA device 5, and extract the information of the edge of the object in the image, including the position and angle. The principle of visual edge capture is to detect the edge based on the grayscale change of the image. The present invention uses the edge detection operator in the multi-level edge detection algorithm to extract the edge of the object; the goal is to find an optimal edge detection solution or find the position with the strongest grayscale intensity change in an image. Achieving optimal edge detection improves the low error rate, high positioning and minimum response performance of the present invention.

[0047] Based on the ideal embodiment of the present invention, through the above description, relevant personnel can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the content in the specification, but must be determined according to the scope of the claims.

Claims

1. A silicon photonic FA alignment device, characterized in that: It comprises a product stage mechanism (4) for receiving and positioning a product component (43), wherein a UV curing mechanism (2) and an alignment structure (3) are arranged on the periphery of the product stage mechanism (4); The product carrier mechanism (4) comprises a carrier leveling mechanism (41) and a carrier (42) arranged on the carrier leveling mechanism (41), and the product component (43) is positioned on the carrier (42); The alignment structure (3) comprises an alignment posture adjustment platform (31), a clamping drive mechanism (32) is arranged on the alignment posture adjustment platform (31), and a clamping mechanism (33) is arranged on the clamping drive mechanism (32); The UV curing mechanism (2) comprises a displacement driving mechanism (21), and a UV curing component (22) and an image detection auxiliary mechanism (23) arranged on the displacement driving mechanism (21); The clamping mechanism (33) obtains the FA device (5), and can adjust the position of the FA device by displacement relative to the product component (43) on the carrier (42) through the clamping drive mechanism (32) and the clamping mechanism (33), so as to achieve alignment between the FA device and the product component (43), and achieve position verification and curing assembly through the UV curing component (22) and the image detection auxiliary mechanism (23) in the UV curing mechanism (2).

2. The silicon photonic FA alignment device according to claim 1, characterized in that: The displacement drive mechanism (21) comprises an XY displacement mechanism and a longitudinal adjustment mechanism arranged on the XY displacement mechanism; the longitudinal adjustment mechanism is provided with a pair of angle adjustment brackets arranged opposite to each other, and the angle adjustment brackets are provided with a pair of UV curing mechanisms (2) arranged opposite to each other.

3. The silicon photonic FA alignment device according to claim 2, characterized in that: An image detection auxiliary mechanism (23) is provided on the displacement drive mechanism (21), and a detection sampling port of the image detection auxiliary mechanism (23) corresponds to the carrier (42) in the product carrier mechanism (4).

4. The silicon photonic FA alignment device according to claim 3, characterized in that: The clamping mechanism (33) comprises an adsorption head (332), and a pair of clamping claws (331) arranged at intervals on both sides of the adsorption head (332); the clamping claws (331) are connected to the alignment posture adjustment platform (31) through the clamping drive mechanism (32), and the clamping drive mechanism (32) drives the clamping claws (331) on both sides of the adsorption head (332) to open and close relative to the adsorption head (332).

5. A method for aligning a silicon photonic FA alignment device, characterized in that: The method is implemented by using the silicon photonic FA alignment device according to any one of claims 1 to 4, comprising the following steps: The initial position between the FA device (5) acquired by the clamping mechanism (33) in the alignment structure (3) and the product component (43) carried by the product stage mechanism (4) is acquired by an image detection auxiliary mechanism (23) arranged on the displacement driving mechanism (21) in the UV curing mechanism (2); Based on the position image between the FA device (5) and the product component (43) obtained by the image detection auxiliary mechanism (23), the relative position relationship is obtained by an image processing algorithm, and the compensation parameters are obtained by using an alignment algorithm based on the relative position relationship. The platform leveling mechanism (41) in the product platform mechanism (4) is driven according to the supplementary parameters to adjust the posture of the product component (43) supported and positioned on the platform (42), and the posture of the FA device (5) is adjusted by adjusting the positioning posture of the platform (31) of the alignment structure (3), so as to achieve alignment between the product component (43) and the FA device (5), so that the product component (43) and the FA device (5) meet the pre-set alignment parameters.

6. The alignment method of silicon photonic FA alignment equipment according to claim 5, characterized in that: After dispensing, the aligned product component (43) and the FA device (5) are UV-cured by the UV curing component (22) of the UV curing mechanism (2).

7. The alignment method of silicon photonic FA alignment equipment according to claim 6, characterized in that: The image processing algorithm comprises the following steps: A visual edge capture algorithm is used to process and analyze the image of the chip and FA device (5) on the product component (43) in the position image, and information on the object edges of the chip and FA device (5) in the image is extracted, wherein the object edge information includes position information and angle information.

8. The alignment method of silicon photonic FA alignment equipment according to claim 7, characterized in that: The visual edge grabbing algorithm comprises the following steps: Step a1, applying Gaussian filtering to smooth the position image and preliminarily remove noise from the position image; Step a2, finding the gradient of the position image; Step a3, applying a non-maximum suppression algorithm to the position image processed in step a2 based on the gradient of the position image to filter out non-edge pixels, making the blurred boundary clear; and retaining the maximum value of the gradient intensity on the pixel point to further filter out noise; Step a4, applying a double threshold method to determine potential boundaries; In step a5, the boundary is tracked using the turbulence post algorithm. When the weak boundary whose pixel position is connected to the strong boundary is considered to be the boundary, the other weak boundaries are deleted.

9. The alignment method of silicon photonic FA alignment equipment according to claim 8, characterized in that: In step a2, the gradient of the position image is found by calculating the gradient magnitude and direction according to the Sobel operator, finding the gradient of the image, and then calculating the gradient magnitude and direction; And / or, in step a3, the non-maximum suppression algorithm is: approximate the gradient direction to one of the proposed values, then the gradient direction represents the up, down, left, right and 45-degree directions; compare the gradient strengths of the corresponding pixel points and the pixel points in the positive and negative directions of the gradient, and retain the pixel point when the gradient strength is the largest, otherwise suppress it; And / or, in step a4, the double threshold method is used to determine the potential boundary: an upper threshold and a lower threshold are set; when the pixel in the image is greater than the upper threshold, it is considered to be a boundary, which is called a strong boundary; when the pixel in the image is less than the lower threshold, it is considered not to be a boundary; when the pixel in the image is between the upper threshold and the lower threshold, it is considered to be a candidate, which is called a weak boundary; And / or, the alignment algorithm includes generalized polynomial fitting using the Bisquare method, and the optimal position is extracted after polynomial fitting is performed on discrete data.

10. The alignment method of silicon photonic FA alignment equipment according to claim 6, characterized in that: The alignment algorithm comprises the following steps: adjusting the angle between the chip and the FA device (5) on the product component (43): obtaining the edge angle x of the chip and the edge angle y of the FA device (5) by the visual edge-grabbing algorithm in the second embodiment, and then calculating the number of rotation axis pulses z by the formula z=(xy)×m; wherein m is the coefficient of the angle-to-pulse number, and after the rotation axis rotates by z pulses, the edges of the chip and the FA device (5) can remain parallel; The alignment algorithm is used to analyze the coordinate axis position corresponding to the peak of the curve during the alignment process, and the stepper axis is controlled by the motion control card to align the optical paths of the chip and the FA device (5) to achieve the maximum optical power value.

Images