A packaging method for vertical cavity surface emitting laser

Abstract

The present invention relates to the field of laser packaging technology, and more specifically, to a packaging method for a vertical cavity surface emitting laser (VCSEL). The method comprises: prior to TO packaging, obtaining images of a tube socket at several Z-axis distances; analyzing the tube socket image blur at each distance, and determining the preferred luminous point center position of the tube socket image at each distance; determining the tilt of the TO cap based on the change in the preferred luminous point center position in the tube socket images at adjacent distances, and using the tilt to correct the coaxial offset of the tube socket; and performing TO packaging based on the alignment result of the TO tube socket and the TO cap after the coaxial offset correction. The present invention significantly enhances the coupling efficiency and packaging consistency of TO devices by overcoming the defect of offset confirmation of the reflection center position caused by image blur.

Description

Technical Field

[0001] The present invention relates to the technical field of laser packaging, and in particular to a packaging method for a vertical cavity surface emitting laser. Background Art

[0002] With the rapid development of optical communications, LiDAR, 3D sensing, and other fields, vertical-cavity surface-emitting lasers (VCSELs) have become one of the most widely used light sources in optoelectronic devices due to their low power consumption, high modulation rate, and array integration capabilities. Driven by the trend towards high-performance, high-density packaging, TO (Transistor Outline) packaging continues to play a key role in VCSEL devices due to its excellent heat dissipation, mechanical stability, and compatibility with optical fibers, lenses, and other components. Existing TO packaging methods generally use monocular vision to acquire images of the VCSEL device's lens and chip, and employ traditional image processing algorithms (such as Hough circle detection) to determine the relative position of the chip and lens, and to analyze and adjust coaxial deviations.

[0003] During the laser packaging process, precise alignment between the TO package lens and the chip is a critical step in ensuring stable optoelectronic performance. However, due to slight differences in focal length, spherical aberration, and other aspects of the non-standardized lenses used in TO caps, and the possibility of tilt in the TO socket during alignment, the captured image of the luminous point may have blurred edges, resulting in errors in determining the center position of the luminous point, reducing the accuracy of coaxial alignment and thus affecting the packaging effect of the vertical cavity surface emitting laser. Summary of the Invention

[0004] In order to solve the technical problem that the coaxial alignment accuracy in the existing packaging process is low, resulting in poor packaging effect of vertical cavity surface emitting lasers, the purpose of the present invention is to provide a packaging method for vertical cavity surface emitting lasers. The technical solution adopted is as follows:

[0005] An embodiment of the present invention provides a method for packaging a vertical cavity surface emitting laser, the method comprising the following steps:

[0006] Before the vertical cavity surface emitting laser is packaged in a TO package, images of the tube base at several distances are obtained; wherein the distance is the Z-axis distance between the TO tube base and the TO tube cap;

[0007] Analyze the blur of the tube base image at each distance and determine the optimal luminous point center position of the tube base image at each distance;

[0008] Determining the tilt of the TO cap according to the change in the center position of the preferred luminous point in the tube socket image at adjacent distances, and using the tilt to correct the coaxial offset of the tube socket;

[0009] Based on the alignment result of the TO tube holder and the TO tube cap after the coaxial offset correction, TO packaging is performed.

[0010] Furthermore, analyzing the blurring of the tube socket image at each distance and determining the optimal luminous point center position of the tube socket image at each distance includes:

[0011] For the tube base image at each distance, edge detection is performed on the tube base image to obtain various edge connected domains, and the blur degree of the tube base image is determined based on the various edge connected domains;

[0012] Determining the gradient correction degree of each edge-connected domain according to the blur degree of the tube seat image and the predicted center position of the edge-connected domain;

[0013] Correcting the original gradient value of each pixel in the tube base image using the gradient correction degree to obtain a corrected gradient value of each pixel;

[0014] The optimal luminous point center position of the tube base image is determined according to the corrected gradient value of each pixel point in the tube base image.

[0015] Furthermore, determining the blur degree of the tube socket image based on each edge connected domain includes:

[0016] Performing Hough circle detection on the tube seat image to obtain a circle under each center position in the tube seat image;

[0017] Determine the degree of diffuse circle of each edge connected domain based on the number of intersections between each edge connected domain and all circles at the center of each circle, the number of edge pixels corresponding to each edge connected domain, and the gradient value of each edge pixel; wherein the degree of diffuse circle indicates the possibility that the edge connected domain belongs to the edge of the diffuse circle;

[0018] The number of edge pixels corresponding to each edge connected domain is used to perform weighted sum analysis on the degree of the circle of confusion to determine the blur degree of the tube seat image.

[0019] Furthermore, the step of determining the degree of the circle of confusion of each edge connected domain according to the number of intersections between each edge connected domain and the circles at all circle center positions, the number of edge pixels corresponding to each edge connected domain, and the gradient value of each edge pixel includes:

[0020] For each edge connected region, selecting a maximum number of intersections from all numbers of intersections corresponding to the edge connected region, and determining a ratio of the maximum number of intersections to the number of edge pixels;

[0021] Calculating the average value of all gradient values ​​corresponding to the edge connected domain, and performing negative correlation processing on the average value of all gradient values ​​to obtain a negative correlation value;

[0022] The degree of the circle of confusion of the edge connected domain is determined by combining the ratio and the negative correlation value corresponding to the edge connected domain.

[0023] Furthermore, the step of determining the gradient correction degree of each edge-connected domain according to the blur degree of the tube seat image and the predicted center position of the edge-connected domain includes:

[0024] Determine the center position of the circle with the largest number of intersections with the target edge connected region as the predicted center position of the target edge connected region;

[0025] determining a gradient correction degree of the target edge connected domain based on a first distance between a predicted center position of the target edge connected domain and predicted center positions of each comparison edge connected domain, a blur degree of the tube seat image, and a degree of a circle of confusion of the target edge connected domain;

[0026] The target edge connected domain is any edge connected domain in the tube seat image, and the comparison edge connected domain is an edge connected domain in the tube seat image other than the target edge connected domain; the first distance, the blur degree, and the degree of the circle of confusion are all proportional to the degree of gradient correction.

[0027] Furthermore, the method of correcting the original gradient value of each pixel in the tube base image by using the gradient correction degree to obtain the corrected gradient value of each pixel includes:

[0028] For each pixel point in the tube base image, determining a second distance between the pixel point and each edge-connected domain;

[0029] Calculating the ratio of the gradient correction degree of each edge connected domain to the second distance to determine the correction weight of the pixel point;

[0030] The original gradient value of the pixel point is corrected using the correction weight to obtain the corrected gradient value of the pixel point.

[0031] Furthermore, determining the optimal luminous point center position of the tube base image according to the corrected gradient value of each pixel point in the tube base image includes:

[0032] Performing Hough circle detection on the tube base image according to the corrected gradient value of each pixel in the tube base image to obtain each new circle center position;

[0033] The average coordinate position of all the new circle center positions is used as the optimal luminous point center position of the tube base image.

[0034] Furthermore, the determining of the tilt degree of the TO cap according to the change of the center position of the preferred luminous point in the tube base image at adjacent distances includes:

[0035] Obtaining each displacement vector according to the difference between the center positions of the preferred luminous points in the tube base images at adjacent distances;

[0036] The movement of the center of the preferred light-emitting point along one direction is analyzed according to the displacement vectors to determine the degree of inclination of the TO cap.

[0037] Furthermore, analyzing the movement of the center of the preferred light-emitting point along one direction according to each displacement vector to determine the tilt degree of the TO cap includes:

[0038] Determine the modulus of each displacement vector, and use the accumulated value of the modulus of all displacement vectors as a first tilt factor;

[0039] determining a standard deviation of all displacement vectors, and using a negative correlation value of the standard deviation as a second tilt factor;

[0040] The first tilt factor and the second tilt factor are combined to obtain the tilt degree of the TO cap.

[0041] Furthermore, the method of correcting the coaxial offset of the tube seat by using the tilt degree includes:

[0042] Setting a tilt threshold, when the tilt degree is greater than the tilt threshold, using the adjusted tilt angle as an adjustment target, and performing corrections through PID adjustment until the tilt degree is no greater than the tilt threshold;

[0043] When the tilt degree is not greater than the tilt threshold, the coaxial offset is corrected by the position difference between the center point of the rectangular chip and the center point of the reflection until the position difference is within the preset error range; wherein, the center point of the rectangular chip is obtained through the chip edge pixel point.

[0044] The present invention has the following beneficial effects:

[0045] The present invention provides a packaging method for vertical-cavity surface-emitting lasers (VCSELs). This method acquires images at multiple Z-axis heights before TO packaging, analyzes the image blur and luminous point displacement, and accurately quantifies the tilt angle of the TO socket. The tilt is then used to correct the socket's coaxial offset. Based on the alignment results of the TO socket and TO cap after the coaxial offset correction, TO packaging is performed. Compared to traditional alignment methods that rely solely on single-frame images, this method can effectively compensate for imaging offsets caused by lens focal length differences or assembly errors, improve the stability and accuracy of optical axis alignment, and thus significantly enhance the coupling efficiency and packaging consistency of TO devices, meeting the stringent requirements for device photoelectric conversion efficiency and reliability in the field of high-performance optical communications. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 This is a flow chart of a method for packaging a vertical cavity surface emitting laser according to the present invention;

[0048] Figure 2 This is an example of the luminous point image of the TO socket chip;

[0049] Figure 3 A simplified visual layout diagram of the TO alignment process;

[0050] Figure 4 Flowchart for implementing step S2 in an embodiment of the present invention;

[0051] Figure 5 Schematic diagram of edge detection results of a tube socket image in an embodiment of the present invention;

[0052] Figure 6 Schematic diagram of Hough circle detection results achieved by original gradient values ​​in an embodiment of the present invention;

[0053] Figure 7 Schematic diagram of Hough circle detection results achieved by modifying the gradient value in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] To further illustrate the technical means and effects employed by the present invention to achieve its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementations, structures, features, and effects of the technical solutions proposed by the present invention. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0055] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0056] The application scenarios targeted by the present invention may be:

[0057] During the TO packaging process of vertical-cavity surface-emitting lasers (VCSELs), optical differences (focal length, spherical aberration) in non-standardized lenses and defocused imaging caused by tube base tilt prevent traditional image algorithms (such as Hough circle detection) from accurately identifying the center of the luminous point. This results in reduced coaxial alignment accuracy and poor laser packaging performance. To improve image clarity and enhance alignment accuracy, a dynamic focusing process based on image clarity feedback, combined with refined image analysis, provides more reliable coaxial deviation information, thereby improving laser packaging performance.

[0058] One embodiment of the present invention provides a method for packaging a vertical cavity surface emitting laser. Figure 1 As shown, the following steps are included:

[0059] S1, before the vertical cavity surface emitting laser is packaged in TO packaging, the tube base images at several distances are obtained.

[0060] Here, the distance is the Z-axis distance between the TO socket and the TO cap. The socket image specifically refers to the luminous point image of the TO socket chip. The example of the luminous point image of the TO socket chip is shown in the figure below. Figure 2 shown.

[0061] During the TO packaging process of VCSEL optical devices, the TO tube holder with the light-emitting chip and the TO tube cap containing the optical lens need to be aligned and packaged with high precision. However, due to the possibility of blurred imaging due to errors in the lens structure and assembly, it is necessary to collect tube holder images at different Z-axis distances before TO packaging. This is to facilitate the subsequent analysis of image clarity and light-emitting point deviation, and ultimately determine the optimal alignment position before packaging.

[0062] The emission path of the light source is merged with the imaging path of the camera. The light from the light source is vertically incident on the TO lens through the beam splitter prism. The light is focused by the lens of the TO tube cap and hits the tube socket chip. The chip surface will reflect or form a clear image point, so there is a chip light point in the collected tube socket image.

[0063] Specifically, the TO cap and TO socket are first secured to a fixture using an adsorption system, initially maintaining their optical axes roughly concentric. The adsorption system, as used here, refers to a system that utilizes adsorption force (typically vacuum or other mechanical adsorption methods) to secure the TO cap and socket to the fixture. The initial Z-axis spacing between the TO cap and socket can be set to approximately 0.5 mm. Starting from this initial spacing, the Z-axis distance between the TO cap and socket is gradually adjusted in fixed steps, pausing approximately 100 ms after each adjustment to capture a single socket image. The fixed step size can be set to 20 μm, and the number of distance adjustments can be set to 20, meaning 20 socket images are captured, covering the 0.1–0.5 mm range.

[0064] It should be noted that the reason why the tube base and the tube cap are coaxial is as follows: When the axes of the tube cap and the tube base are aligned by industrial addition, the visual layout diagram of the TO alignment process is as follows Figure 3 As shown in the figure, there is an optical lens above the tube cap. The light emitted by the VCSEL needs to be collimated or focused by the lens to ensure that the light can propagate in the correct direction and shape. If the lens and the VCSEL are not coaxial, the propagation direction and intensity distribution of the light may be affected, resulting in a decrease in beam quality. Therefore, the VCSEL on the tube base should be aligned with the axis of the tube cap.

[0065] Thus, this embodiment obtains the tube seat image at each distance for analyzing the coaxial offset between the tube seat and the tube cap.

[0066] S2, analyzing the blurring of the tube socket image at each distance, and determining the optimal luminous point center position of the tube socket image at each distance.

[0067] Here, the preferred luminous point center position refers to the luminous point center position of the luminous point image of the TO tube socket chip after overcoming the blur effect. By obtaining a more accurate luminous point center position, the coaxial alignment accuracy of the tube cap and the tube socket during the TO packaging process can be ensured to a certain extent.

[0068] Since the lens on the TO cap is usually a convex lens structure, its focal length is affected by factors such as glass material, curvature radius, lens thickness, etc., and lenses from different batches or even the same batch may have slight focal length differences. Therefore, if the imaging surface of the visual system does not fall accurately on the light-emitting area of ​​the VCSEL chip or the focus of the lens, the collected image will be blurred, resulting in unclear edges of the light-emitting point, failure of image processing algorithms such as the Hough circle, or increased errors. The center position of the light-emitting point determined based on this will also have deviations. Specifically, Figure 2 In the process, there is a certain coaxial offset between the center of the luminous point and the chip position. The center of the luminous point is usually determined based on the traditional Hough circle detection method. Under the premise that the tube holder image is blurred, the accuracy of the luminous point center position determined by it is low.

[0069] For the tube base image at each distance, it is preferred that the method for determining the center position of the luminous point is consistent. Taking any tube base image as an example, as an exemplary implementation, the above step S2 can be performed by Figure 4 Steps S201 to S204 shown implement:

[0070] S201 , performing edge detection on a tube socket image to obtain edge connected domains, and determining a blur degree of the tube socket image based on the edge connected domains.

[0071] Here, the edge connected domain means that if a group of edge pixels are connected to each other through 8-connectivity or 4-connectivity, then this group of edge pixels constitutes an edge connected domain; the blur degree refers to the imaging blur when the VCSEL chip or its reflection or luminous point does not fall on the focal plane of the imaging system.

[0072] Firstly, the Canny edge detection algorithm is used to perform edge detection on the tube seat image to obtain the edge detection result, and each edge connected domain in the edge detection result is extracted.

[0073] In this embodiment, the edge detection result diagram is as follows: Figure 5 As shown, the implementation process of the Canny edge detection algorithm is a prior art and is not within the scope of protection of the present invention, and will not be elaborated here.

[0074] Secondly, the probability of each edge connected domain being on the same circle and the gradient size of the edge connected domain are analyzed to determine the blur degree of the tube seat image.

[0075] Depend on Figure 5 As can be seen, the chip edge has a significantly different distance from the surrounding area on the tube holder. Therefore, the chip edge often has a large gradient, while the edge of the circle of confusion is relatively blurred. This can lead to inaccurate determination of the chip center and the reflection center. Therefore, it is necessary to determine the degree of blur in the tube holder image.

[0076] When the VCSEL chip or its reflection or luminous point does not fall on the focal plane of the imaging system, the point light source appears as a diffusion circle in the image, that is, a circle of gradually decaying brightness rings, with a bright center, gradually darkening edges, and blurred edges, forming a halo-like effect. Due to the blurred edges of the diffusion circle and the slow grayscale changes, traditional edge detection algorithms have difficulty detecting clear edges. In other words, the extracted edge pixels are sparse and broken, resulting in an insufficient number of "valid voting" points relied on by the Hough algorithm for determining the center position of the luminous point, leading to missed detections, false detections, or fitting offsets. Therefore, before performing Hough circle detection, it is necessary to quantitatively analyze the blur of the tube base image to facilitate the subsequent use of different gradient enhancement methods for tube base images with different blur levels to improve the accuracy of the final determination of the preferred luminous point center position.

[0077] Preferably, the method for determining the blur degree of the tube seat image includes:

[0078] The first step is to perform Hough circle detection on the tube socket image to obtain the circle under each center position in the tube socket image.

[0079] Specifically, Hough circle detection is performed on the edge detection results of the tube socket image to obtain circles with different center positions. The implementation process of Hough circle detection is prior art and is not within the scope of protection of the present invention, so it will not be elaborated here.

[0080] In the second step, the degree of diffuse circle of each edge connected domain is determined according to the number of intersections between each edge connected domain and the circles under all circle center positions, the number of edge pixels corresponding to each edge connected domain, and the gradient value of each edge pixel.

[0081] Here, the degree of the diffuse circle indicates the possibility that the edge connected domain belongs to the diffuse circle edge.

[0082] In this embodiment, the greater the probability that an edge-connected domain lies on the same circle, and the more locally blurred the edge-connected domain is, the more likely it is that the corresponding edge-connected domain belongs to the edge of a diffuse circle. The ratio of the number of intersections between the edge-connected domain and the circle to the number of edge pixels can represent the probability that the edge-connected domain lies on the same circle, while the gradient of the edge-connected domain can represent the degree of local blur in the edge-connected domain. Based on this, the degree of diffuse circle of the edge-connected domain can be determined.

[0083] Preferably, the degree of the circle of confusion of the edge connected domain is determined by:

[0084] In the first sub-step, the maximum number of intersections is selected from all the numbers of intersections corresponding to the edge connected domain, and the ratio of the maximum number of intersections to the number of edge pixels is determined.

[0085] In the second sub-step, the average value of all gradient values ​​corresponding to the edge connected domain is calculated, and negative correlation processing is performed on the average value of all gradient values ​​to obtain a negative correlation value.

[0086] In the third sub-step, the degree of the circle of confusion of the edge connected domain is determined by combining the ratio and negative correlation value corresponding to the edge connected domain.

[0087] As an example, the calculation formula for the degree of the circle of confusion of the i-th edge connected domain can be:

[0088] Where, Indicates the degree of diffusion circle of the ith edge connected domain, max represents the maximum value function, represents the number of intersections between the circle at the rth circle center position and the i-th edge connected domain, represents the center set of the tube seat image belonging to the i-th edge connected domain in the Hough space, Represents the number of edge pixels in the i-th edge connected domain, Represents the ratio of the maximum number of intersection points to the number of edge pixels, Represents the average value of all gradient values ​​corresponding to the i-th edge connected domain, The negative correlation value representing the average of all gradient values.

[0089] In the calculation formula of the degree of confusion circle, Refers to the probability that the i-th edge-connected domain is on the same circle, Refers to the degree of blur of the i-th edge connected domain. The smaller the gradient, the greater the blur of the edge connected domain. When the i-th edge connected domain is locally blurred and has a high probability of being on the same circle, the probability that the i-th edge connected domain is the edge of a diffuse circle is high.

[0090] With reference to the method for determining the degree of the circle of confusion of the i-th edge-connected domain, the degree of the circle of confusion of each edge-connected domain in the tube seat image can be obtained.

[0091] In the third step, the number of edge pixels corresponding to each edge connected domain is used to perform weighted sum analysis on the degree of confusion circle to determine the blur degree of the tube seat image.

[0092] If the tube socket image contains multiple edge-connected domains with high circular diffusion, the tube socket image will be less clear and more blurred, requiring subsequent Hough circle detection parameter adjustment, i.e., determining a corrected gradient value. The greater the number of edge pixels in an edge-connected domain, the greater the confidence level in the circular diffusion of that edge-connected domain. Therefore, the number of edge pixels can be used to weight the diffusion circle degree, which helps improve the numerical accuracy of the calculated blur degree.

[0093] As an example, the calculation formula for the blur degree of the h-th tube socket image may be:

[0094] Where, Indicates the blur level of the h-th tube base image, norm indicates the linear normalization function, Indicates the number of edge pixels corresponding to the i-th edge connected domain, Indicates the degree of diffusion circle of the i-th edge connected domain, , I represents the number of edge-connected domains in the h-th tube socket image.

[0095] In another example, the calculation formula for the blur degree of the h-th tube socket image may also be:

[0096] ;

[0097] S202 , determining the gradient correction degree of each edge-connected domain according to the blur degree of the tube seat image and the predicted center position of the edge-connected domain.

[0098] Here, the gradient correction degree refers to the degree to which the gradient value of the edge connected component needs to be corrected, which is determined by taking into account the image blur and image formation characteristics.

[0099] The degree of blur of the tube socket image varies at different Z-axis distances. For tube socket images with a larger degree of blur, they are more affected by the confusion circle when determining the reflection center. Therefore, the first gradient correction factor is determined based on the degree of blur of the tube socket image and the degree of confusion circle of the edge connected domain. For edge pixels on the edge connected domain, the tube socket image to which they belong is affected by the light reflected on the tube socket after passing through the lens. The result of Hough circle detection on the tube socket image should be multiple concentric circles. Therefore, when performing gradient enhancement on a single edge connected domain, it is also necessary to consider whether the center positions of the circles corresponding to each edge connected domain are the same or close, that is, to determine the second gradient correction factor.

[0100] Preferably, the target edge connected domain is any edge connected domain in the tube base image, and determining the gradient correction degree of the edge connected domain using the target edge connected domain as an example includes:

[0101] The first step is to determine the center position of the circle with the largest number of intersections with the target edge connected domain, which is used as the predicted center position of the target edge connected domain.

[0102] Specifically, first obtain the circles at each center position corresponding to the target edge connected domain, count the intersections between the target edge connected domain and the circles at each center position, and then determine the center position with the largest number of intersections as the predicted center position of the target edge connected domain.

[0103] As an example, the calculation formula for the predicted circle center position corresponding to the i-th edge connected region can be:

[0104] Where, Indicates the predicted center position of the i-th edge connected domain, argmax represents the maximum point function, represents the number of intersections between the circle at the rth circle center position and the i-th edge connected domain, represents the circle center set of the tube base image to which the ith edge-connected domain belongs in the Hough space, and the ith edge-connected domain may be the target edge-connected domain.

[0105] It should be noted that the principle of determining the predicted center position is: a center positioning algorithm based on geometric probability voting, the core of which is to infer the optimal center by counting the geometric consistency between edge points and candidate circles, which facilitates the subsequent determination of the first distance.

[0106] In the second step, the degree of gradient correction of the target edge connected domain is determined based on the first distance between the predicted center position of the target edge connected domain and the predicted center positions of each comparison edge connected domain, the blur degree of the tube seat image, and the degree of the diffusion circle of the target edge connected domain.

[0107] Here, the contrast edge connected domain is the edge connected domain in the tube seat image except the target edge connected domain; the first distance, blur degree, and confusion circle degree are all proportional to the gradient correction degree, that is, the greater the first distance, blur degree, and confusion circle degree corresponding to the target edge connected domain, the greater the gradient correction degree of the target edge connected domain.

[0108] In the first sub-step, a first gradient correction factor is determined according to the blur degree of the tube seat image and the degree of the circle of confusion of the target edge connected domain.

[0109] Specifically, the product of the blur degree and the circle of confusion degree is calculated, and the product is normalized using a linear normalization function to obtain a first gradient correction factor.

[0110] In the second sub-step, a second gradient correction factor is determined according to a first distance between the predicted center position of the target edge connected region and the predicted center positions of each comparison edge connected region.

[0111] Specifically, the accumulated value or average value of all the first distances is calculated, and the accumulated value or average value is negatively correlated to obtain a negative correlation value as the second gradient correction factor.

[0112] In the third sub-step, the first gradient correction factor and the second gradient correction factor are combined to obtain the gradient correction degree of the target edge connected domain.

[0113] Specifically, the product of the first gradient correction factor representing the blurring of the image and the target edge connected domain and the second gradient correction factor representing the probability that the target edge connected domain is a multi-circle diffuse circle edge can be used as the gradient correction degree of the target edge connected domain.

[0114] In this embodiment, when the distance between the predicted circle center position and the predicted circle centers of multiple edge connected domains is small, it indicates that the probability that the i-th edge connected domain is a multi-circle diffuse circle edge is higher, and the degree of alignment correction should be higher. Then, the degree of gradient correction of the i-th edge connected domain is determined by combining the degree of blur and the degree of diffuse circle.

[0115] As an example, the calculation formula for the gradient correction degree of the i-th edge connected domain can be:

[0116] Where, It represents the gradient correction degree of the ith edge connected domain in the hth tube seat image, norm represents the linear normalization function, Indicates the blur level of the h-th tube base image, Indicates the degree of diffusion circle of the i-th edge connected domain, represents the number of contrast edge-connected domains in the h-th tube base image except the i-th edge-connected domain, Indicates the predicted center position corresponding to the i-th edge connected domain, Indicates the predicted circle center position corresponding to the j-th contrast edge connected domain, Indicates the predicted center position and The first distance between.

[0117] In the calculation formula of the gradient correction degree, It can also be expressed as the set of the first K predicted circle center positions with the smallest distance between them, where the empirical value of K can be 20; Represents the first gradient correction factor. In order to avoid dimension problems, it is necessary to Perform normalization processing, Indicates the blur of the entire image, which can reflect the problem of image shooting and analyze the degree of diffusion circle of the edge connected domain The purpose is to distinguish the edge of the diffusion circle from the edge of the chip; Represents the second gradient correction factor, which is used to analyze whether the center positions of the circles corresponding to the i-th edge connected domain are the same or close. In general, There is no possibility of zero. If there is an extreme case, add a non-zero constant, such as 0.001, to the denominator of the fraction.

[0118] S203, correcting the original gradient value of each pixel in the tube base image using the gradient correction degree to obtain a corrected gradient value of each pixel.

[0119] Preferably, taking any pixel point as an example, determining the corrected gradient value of the pixel point includes:

[0120] The first step is to determine the second distance between the pixel and each edge connected domain.

[0121] Specifically, the coordinate position of the pixel point in the image and the coordinate position of the geometric center of the edge connected domain in the image are obtained, and the Euclidean distance between the two points is calculated based on the two coordinate positions as the second distance.

[0122] The second step is to calculate the ratio of the gradient correction degree of each edge connected domain to the second distance, and determine the correction weight of the pixel point.

[0123] Specifically, each ratio is normalized to obtain each normalized value, and the accumulated value of all normalized values ​​is used as the correction weight of the pixel point.

[0124] The third step is to use the correction weight to correct the original gradient value of the pixel point to obtain the corrected gradient value of the pixel point.

[0125] As an example, the calculation formula for the corrected gradient value of the u-th pixel in the h-th tube base image can be:

[0126] Where, represents the corrected gradient value of the u-th pixel in the h-th tube base image, N represents the number of edge connected domains in the h-th tube base image, norm represents the linear normalization function, Indicates the degree of gradient correction of the i-th edge-connected domain in the h-th tube base image, It represents the second distance between the u-th pixel point and the i-th edge-connected domain in the h-th tube base image. The second distance is generally not zero. If there is an extreme case, a non-zero constant is added to the denominator of the fraction, such as 0.001. represents the correction weight of the u-th pixel in the h-th tube base image, Represents the original gradient value of the u-th pixel in the h-th tube base image.

[0127] With reference to the above method for determining the corrected gradient value of the u-th pixel, the corrected gradient value of each pixel can be obtained.

[0128] S204: Determine the optimal center position of the luminous point of the tube socket image according to the corrected gradient value of each pixel in the tube socket image.

[0129] In the first step, Hough circle detection is performed on the tube base image according to the corrected gradient value of each pixel in the tube base image to obtain the new center position of each circle.

[0130] In the second step, the average coordinate position of all new circle center positions is used as the optimal luminous point center position of the tube base image.

[0131] In this embodiment, when the Hough circle detection is performed on the tube base image using the corrected gradient value, if there is only one new circle center position, it is used as the preferred luminous point center position; if there are no less than two new circle center positions, the average value of all new circle center positions is used as the preferred luminous point center position.

[0132] Among them, the schematic diagram of the Hough circle detection result realized by the original gradient value is as follows Figure 6 As shown, the schematic diagram of the Hough circle detection result achieved by modifying the gradient value is as follows Figure 7 shown.

[0133] Thus, this embodiment obtains a more accurate and preferred center position of the light-emitting point in the tube base image at each distance.

[0134] S3, determining the tilt degree of the TO tube cap according to the change of the center position of the preferred luminous point in the tube socket image at adjacent distances, and correcting the coaxial offset of the tube socket using the tilt degree.

[0135] Here, the tilt degree of the TO tube cap has important reference value for subsequent tilt compensation or fixture precision adjustment, which can greatly improve the packaging accuracy and consistency.

[0136] During the packaging process, the TO socket may be slightly tilted, causing its light-emitting chip to be not completely parallel to the optical axis of the TO cap. Under this premise, even if the central axis of the lens is aligned with the light-emitting point at a certain height, the projected position of the light-emitting point on the image will still shift regularly as the Z axis moves. Therefore, by analyzing the changing trajectory of the light-emitting point in the socket image at multiple Z axis distances, we can analyze whether the light-emitting point has a displacement trend on the ZX / ZY axis and thus determine whether the TO socket is tilted.

[0137] First, the tilt degree of the TO cap is determined according to the change of the center position of the preferred luminous point in the tube base image at adjacent distances.

[0138] At different Z-axis distances, the corresponding optimal luminous point center position may fluctuate within a small range. However, when the optimal luminous point center moves in the same direction, the corresponding TO cap tilts significantly. Under this premise, even if the coaxial offset between the VCSEL in the socket and the luminous center point is calculated, the cap and socket cannot be properly aligned. Therefore, before performing coaxial offset correction, the tilt of the TO cap should be determined first.

[0139] In the first step, each displacement vector is obtained according to the difference between the center positions of the preferred luminous points in the tube base image at adjacent distances.

[0140] In the second step, the movement of the center of the preferred light-emitting point along one direction is analyzed according to each displacement vector to determine the tilt degree of the TO cap.

[0141] Furthermore, determining the tilt degree of the TO cap includes:

[0142] In the first sub-step, the modulus of each displacement vector is determined, and the accumulated value of the modulus of all displacement vectors is used as the first tilt factor.

[0143] The second sub-step is to determine the standard deviation of all displacement vectors, and use the negative correlation value of the standard deviation as the second tilt factor.

[0144] In the third sub-step, the first tilt factor and the second tilt factor are combined to obtain the tilt degree of the TO cap.

[0145] As an example, the calculation formula for the inclination of the TO cap can be:

[0146] Where, Indicates the tilt degree of the TO tube cap, H indicates the number of tube seat images, Indicates the center position of the preferred luminous point in the h-th tube base image, norm represents the linear normalization function, and the normalization process facilitates the subsequent coaxial offset correction. represents the center position of the preferred luminous point in the h-1th tube base image, Indicates from point to , Represents the vector norm, Indicates the standard deviation function. Generally speaking, There is no possibility of zero. If there is an extreme case, add a non-zero constant, such as 0.001, to the denominator of the fraction.

[0147] In the calculation formula of the tilt degree, , the denominator of the fraction Represents the regularity characteristics of each displacement vector, that is, the possibility of belonging to the same direction, the numerator of the fraction It represents the displacement degree of each displacement vector; It can also show the difference in the center point of the reflection at different Z-axis distances, that is, the stability of the difference in the center point at adjacent Z-axis distances when moving in one direction. The smaller the standard deviation of all displacement vectors, the greater the tilt of the TO cap.

[0148] Secondly, the coaxial offset of the tube seat is corrected using the degree of inclination.

[0149] It should be noted that when the degree of tilt is small, the offset can be corrected by the positional relationship between the chip center point and the luminous center point. Otherwise, the tilt angle needs to be adjusted first and then the offset correction is performed.

[0150] The first step is to set the tilt threshold. When the tilt degree is greater than the tilt threshold, the adjusted tilt angle is used as the adjustment target and correction is performed through the PID adjustment method until the tilt degree is no greater than the tilt threshold.

[0151] In this embodiment, the tilt threshold is set to 0.85. When the TO cap is clamped, the tilt angle is first corrected. Here, the correction is made by PID adjustment. The tilt angle after each adjustment is used as the adjustment target. The tilt angle is adjusted by changing the angle at which the alignment component holds the TO cap to improve the accuracy of the coaxial connection. The implementation process of the PID adjustment method is prior art and is not within the scope of protection of the present invention, and will not be elaborated here.

[0152] In the second step, when the tilt degree is not greater than the tilt threshold, the coaxial offset is corrected according to the position difference between the center point of the rectangular chip and the center point of the reflection until the position difference is within the preset error range.

[0153] In this embodiment, when When the tube base image at each distance is obtained, the tube base image with the smallest blur is selected to determine the reflection center point. Based on the traditional edge detection (under high threshold), the chip edge pixel points are obtained to determine the center point of the rectangular chip. The coaxial offset is corrected based on the position difference between the center point of the rectangular chip and the reflection center point.

[0154] Furthermore, the coaxial offset correction is performed based on the position difference between the center point of the rectangular chip and the center point of the reflection, including:

[0155] To ensure that the center of the rectangular chip coincides with the center of the reflector, improving the laser heater packaging effect, the motors of the electric X and Y axis modules are controlled to make corrections. When the position difference between the two center points is relatively small and within the preset error range, the coaxial offset correction is stopped. The preset error range can be set to 5-20μm, which can be set by the implementer according to actual conditions and is not specifically limited here.

[0156] S4, performing TO packaging based on the alignment result of the TO tube holder and the TO tube cap after the coaxial offset is corrected.

[0157] After the system confirms that the optical axes are coaxial and the tilt meets process tolerances, it drives the Z-axis module to slowly descend, gradually approaching the tube cap and tube socket. When the two reach the set welding distance, the electric welding / laser welding mechanism is activated to complete the sealing and packaging of the TO device. After packaging, the system performs rapid optical inspection or functional lighting testing to confirm the package consistency and optoelectronic performance. The set welding distance can be 0 mm to 0.05 mm, which is not specified here.

[0158] The present invention provides a packaging method for vertical cavity surface emitting lasers, which obtains images at multiple Z-axis heights before TO packaging, analyzes the image blur and luminous point displacement, and accurately determines the axis deviation and tilt angle between the TO tube cap lens and the tube base chip. After completing the tilt correction, the system drives the alignment platform to accurately adjust the tube base position based on the X, Y offset information extracted from the image, and performs the packaging welding operation after ensuring that the coaxial accuracy meets the standard. Compared with the traditional alignment method that only relies on a single frame image, the present invention can effectively compensate for the imaging offset caused by lens focal length differences or assembly errors, improve the stability and accuracy of optical axis alignment, thereby significantly enhancing the coupling efficiency and packaging consistency of the TO device, and adapting to the stringent requirements of the high-performance optical communication field on the photoelectric conversion efficiency and reliability of the device.

[0159] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A packaging method for a vertical cavity surface emitting laser, characterized in that: The following steps are involved: Before the vertical cavity surface emitting laser is packaged in a TO package, images of the tube base at several distances are obtained; wherein the distance is the Z-axis distance between the TO tube base and the TO tube cap; Analyze the blur of the tube base image at each distance and determine the optimal luminous point center position of the tube base image at each distance; Determining the tilt of the TO cap according to the change in the center position of the preferred luminous point in the tube socket image at adjacent distances, and using the tilt to correct the coaxial offset of the tube socket; Based on the alignment result of the TO tube holder and the TO tube cap after the coaxial offset correction, TO packaging is performed; Analyzing the blur of the tube socket image at each distance and determining the optimal luminous point center position of the tube socket image at each distance includes: For the tube base image at each distance, edge detection is performed on the tube base image to obtain various edge connected domains, and the blur degree of the tube base image is determined based on the various edge connected domains; Determining the gradient correction degree of each edge-connected domain according to the blur degree of the tube seat image and the predicted center position of the edge-connected domain; Correcting the original gradient value of each pixel in the tube base image using the gradient correction degree to obtain a corrected gradient value of each pixel; The optimal luminous point center position of the tube base image is determined according to the corrected gradient value of each pixel point in the tube base image.

2. The packaging method of a vertical cavity surface emitting laser according to claim 1, characterized in that: The determining of the blur degree of the tube seat image based on each edge connected domain includes: Performing Hough circle detection on the tube seat image to obtain a circle under each center position in the tube seat image; Determine the degree of diffuse circle of each edge connected domain based on the number of intersections between each edge connected domain and all circles at the center of each circle, the number of edge pixels corresponding to each edge connected domain, and the gradient value of each edge pixel; wherein the degree of diffuse circle indicates the possibility that the edge connected domain belongs to the edge of the diffuse circle; The number of edge pixels corresponding to each edge connected domain is used to perform weighted sum analysis on the degree of the circle of confusion to determine the blur degree of the tube seat image.

3. The packaging method of a vertical cavity surface emitting laser according to claim 2, characterized in that: The method of determining the degree of diffuse circle of each edge connected domain according to the number of intersections between each edge connected domain and circles at all circle center positions, the number of edge pixels corresponding to each edge connected domain, and the gradient value of each edge pixel, includes: For each edge connected region, selecting a maximum number of intersections from all numbers of intersections corresponding to the edge connected region, and determining a ratio of the maximum number of intersections to the number of edge pixels; Calculating the average value of all gradient values ​​corresponding to the edge connected domain, and performing negative correlation processing on the average value of all gradient values ​​to obtain a negative correlation value; The degree of the circle of confusion of the edge connected domain is determined by combining the ratio and the negative correlation value corresponding to the edge connected domain.

4. The packaging method of a vertical cavity surface emitting laser according to claim 3, characterized in that: The step of determining the gradient correction degree of each edge-connected domain according to the blur degree of the tube seat image and the predicted center position of the edge-connected domain comprises: Determine the center position of the circle with the largest number of intersections with the target edge connected region as the predicted center position of the target edge connected region; determining a gradient correction degree of the target edge connected domain based on a first distance between a predicted center position of the target edge connected domain and predicted center positions of each comparison edge connected domain, a blur degree of the tube seat image, and a degree of a circle of confusion of the target edge connected domain; The target edge connected domain is any edge connected domain in the tube seat image, and the comparison edge connected domain is an edge connected domain in the tube seat image other than the target edge connected domain; the first distance, the blur degree, and the degree of the circle of confusion are all proportional to the degree of gradient correction.

5. The packaging method of a vertical cavity surface emitting laser according to claim 4, characterized in that: The method of correcting the original gradient value of each pixel in the tube base image by using the gradient correction degree to obtain the corrected gradient value of each pixel includes: For each pixel point in the tube base image, determining a second distance between the pixel point and each edge-connected domain; Calculating the ratio of the gradient correction degree of each edge connected domain to the second distance to determine the correction weight of the pixel point; The original gradient value of the pixel point is corrected using the correction weight to obtain the corrected gradient value of the pixel point.

6. The method for packaging a vertical cavity surface emitting laser according to claim 5, wherein: The step of determining the optimal luminous point center position of the tube base image according to the corrected gradient value of each pixel point in the tube base image includes: Performing Hough circle detection on the tube base image according to the corrected gradient value of each pixel in the tube base image to obtain each new circle center position; The average coordinate position of all the new circle center positions is used as the optimal luminous point center position of the tube base image.

7. The packaging method of a vertical cavity surface emitting laser according to claim 1, characterized in that: The determining the tilt degree of the TO cap according to the change of the center position of the preferred luminous point in the tube base image at adjacent distances includes: Obtaining each displacement vector according to the difference between the center positions of the preferred luminous points in the tube base images at adjacent distances; The movement of the center of the preferred light-emitting point along one direction is analyzed according to the displacement vectors to determine the degree of inclination of the TO cap.

8. The method for packaging a vertical cavity surface emitting laser according to claim 7, wherein: Analyzing the movement of the center of the preferred light-emitting point along one direction according to each displacement vector to determine the tilt degree of the TO cap includes: Determine the modulus of each displacement vector, and use the accumulated value of the modulus of all displacement vectors as a first tilt factor; determining a standard deviation of all displacement vectors, and using a negative correlation value of the standard deviation as a second tilt factor; The first tilt factor and the second tilt factor are combined to obtain the tilt degree of the TO cap.

9. The method for packaging a vertical cavity surface emitting laser according to claim 1, wherein: The method of correcting the coaxial offset of the tube seat by utilizing the tilt degree includes: Setting a tilt threshold, when the tilt degree is greater than the tilt threshold, using the adjusted tilt angle as an adjustment target, and performing corrections through PID adjustment until the tilt degree is no greater than the tilt threshold; When the tilt degree is not greater than the tilt threshold, the coaxial offset is corrected by the position difference between the center point of the rectangular chip and the center point of the reflection until the position difference is within the preset error range; wherein, the center point of the rectangular chip is obtained through the chip edge pixel point.

Images