Optical camera system, chip mounter and imaging method thereof

Abstract

The invention discloses an optical camera system, a chip mounter and an imaging method of the chip mounter. An optical camera mechanism is laterally moved into a bonding space through a telescopic mechanism, double cameras are initialized, and an upper top plate and a lower base are positioned to preset positions; synchronous reflection imaging based on a double-optical-surface light splitting layer: a first optical surface reflects a connector image to a first camera, a second optical surface reflects a chip image to a second camera, and a double-target original image set is generated; performing feature extraction on the image set, identifying key feature points of the chip and the connector, and generating a coordinate data set; calculating the position offset of the feature coordinates of the two images, and judging whether the position offset meets a coincidence threshold value or not; if not, the driving mechanism adjusts the position of the bonding space and performs imaging again, and if yes, an alignment signal is output to form a closed-loop control chain; according to the method, double-target real-time capture is realized through synchronous reflection of the light splitting layer, the alignment precision is remarkably improved, and the chip bonding efficiency and yield are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of optical technology, and in particular to an optical imaging system, a patch printer, and an imaging method thereof. Background Technology

[0002] Optical alignment imaging technology is a core component of chip bonding processes, widely used in semiconductor packaging, chip device manufacturing, and 3D integration. As chip dimensions shrink to the nanometer scale and the number of stacked layers increases, the bonding process needs to achieve micrometer-level precision alignment between the chip and bonding components (such as substrates and interposers) while meeting high-throughput production demands. Traditional methods, relying on manual calibration or multi-device collaboration, have become bottlenecks restricting yield and capacity improvement.

[0003] In existing technologies, chip alignment imaging technology mainly adopts time-division multi-target acquisition schemes, such as single-camera mechanical switching: the camera is moved by a translation stage to acquire images of the chip (position A) and the connector (position B) successively, with an interval of ≥0.5 seconds between the two acquisitions; or rotating beam splitter alternating scheme, where a rotating prism guides the optical path in time division, allowing the same camera to alternately acquire images of the two targets. The core defect of this scheme is that the time division operation introduces an uncontrollable timing error chain, resulting in position mismatch at the time of the two image acquisitions.

[0004] Therefore, it is necessary to improve the existing chip alignment imaging technology to solve the technical problem of cumulative positioning error caused by images. Summary of the Invention

[0005] The purpose of this invention is to provide an optical imaging system, a chip mounter, and an imaging method thereof to solve the above-mentioned technical problems.

[0006] To achieve this objective, the present invention adopts the following technical solution: An imaging method for an optical camera system includes the following steps: The optical camera mechanism is driven laterally through the telescopic mechanism into the bonding space between the upper top plate and the lower base, and the camera and optical lens are initialized. At the same time, the upper top plate and the lower base are adjusted to the preset imaging position through the drive mechanism. Based on the beam-splitting characteristics of optical lenses, dual-target reflection imaging is performed synchronously through optical lenses, enabling the camera to capture the reflection images of the chip and the bonding components, thereby generating the original image dataset. Preprocessing and feature extraction operations are performed on the original image dataset to identify key feature points of the chip and the bonding components, and to generate a feature coordinate dataset. Based on the feature coordinate dataset, the positional offset between the two reflection images is calculated, and it is determined whether the preset overlap condition is met, generating positional offset parameters and overlap status flags. If the overlap status flag indicates that the plates are not overlapped, a drive command is generated based on the position offset parameter to control the drive mechanism to adjust the relative position of the upper top plate and the lower base, and then the process returns to acquire the virtual image. If the overlap status flag indicates that the plates are overlapped, an alignment completion signal is output.

[0007] Optionally, the step of laterally driving the optical camera mechanism through the telescopic mechanism into the bonding space between the upper top plate and the lower base, initializing the camera and optical lenses, and simultaneously adjusting the upper top plate and the lower base to a preset imaging position through the drive mechanism specifically includes: The drive path planning of the telescopic mechanism is generated based on the height parameter of the bonding space. The optical camera mechanism is laterally driven by a linear motor to move along the guide rail, and displacement sensor data is collected in real time to correct the position deviation until the central axis of the optical camera mechanism is aligned with the central plane of the bonding space. The camera and light source modules of the optical camera mechanism are activated to collect background light intensity data of the bonding space. The output power of the light source is dynamically adjusted based on the preset light intensity threshold to generate anti-interference environmental parameters. The beam splitter calibration coefficient of the optical lens is loaded into the data processing unit, and the real-time compensation parameters of the optical lens are calculated in combination with the anti-interference environment parameters. The piezoelectric fine-tuner is then controlled to adjust the pitch angle of the optical lens. The encoder of the drive mechanism acquires real-time distance data between the upper top plate and the lower base, compares it with the target distance at the preset imaging position, calculates the position adjustment amount, and drives the upper top plate and the lower base to move towards each other to the target distance, and locks the position of the drive mechanism.

[0008] Optionally, the optical lens includes a lens body, the upper end surface of the lens body is provided with a first optical surface, the lower end surface of the lens body is provided with a second optical surface, and both the first optical surface and the second optical surface are provided with a beam-splitting layer; The first optical surface and the second optical surface are respectively set at a preset angle to the center line of the lens body.

[0009] Optionally, based on the beam-splitting characteristics of the optical lens, the dual-target reflection imaging operation is performed simultaneously through the optical lens, enabling the camera to capture the reflection image of the chip and the reflection image of the bonding component, generating an original image dataset. This specifically includes the following steps: Based on the ambient light intensity sensor readings of the bonding space and the material reflectivity of the chip / bonding component, the incident beam power range of the light source module is calculated, a beam intensity distribution matrix is ​​generated, and the LED array is controlled to project polarized light with a wavelength of 520-650nm into the bonding space at an incident angle of 45° to form an anti-interference imaging light field. S22, the light reflected from the bonding component enters the first optical surface beam splitter at an incident angle θ1=45±0.5°, where the incident light is reflected by the beam splitter to generate the first reflected light, which is transmitted to the first camera; simultaneously, the light reflected from the chip enters the second optical surface beam splitter at an incident angle θ2=45±0.5°, and its reflected light is transmitted to the second camera. The exposure sequence of the first and second cameras is triggered synchronously, and the first and second reflected rays form a positive interactive complementary view on the corresponding imaging plane to generate an initial image pair covering the chip pads and bonding assembly alignment marks.

[0010] Optionally, the step of generating a reflected image covering the chip pads and bonding assembly alignment marks by forming a positive interactive complementary viewing angle on the corresponding imaging plane through the first reflected light and the second reflected light further includes: The real-time compensation parameters of the optical lens are called, and the perspective transformation matrix operation is performed on the initial image pair through the real-time compensation parameters to correct the geometric distortion caused by the fine adjustment of the lens angle and generate a pixel-level aligned calibration image pair. The calibration image pairs are time-stamped and encoded with the real-time compensation parameters of the optical lens, the beam splitter calibration coefficient, and the anti-interference environment parameters to generate an original image dataset including geometric distortion correction information, and stored in a cache queue.

[0011] Optionally, based on the feature coordinate dataset, the positional offset between the reflected virtual image and the refracted virtual image is calculated, specifically including the following steps: The chip feature coordinates and the bonding component feature coordinates in the feature coordinate dataset are mapped to a unified coordinate system with the center of the bonding space as the origin, respectively, to generate the chip standard coordinate set and the bonding component standard coordinate set; Based on the structural similarity between the chip and the bonding component, a dynamic matching weight is defined. The weighted nearest neighbor algorithm is used to perform feature point pairing on the standard coordinate set of the chip and the standard coordinate set of the bonding component to generate a set of matching feature point pairs. Calculate the Euclidean distance and angular deviation of each pair of feature points in the matching feature point pair set, generate a local offset vector set, and remove outlier point pairs that exceed the preset tolerance range; The local offset vector set is optimized by least squares, and the global offset transformation matrix is ​​fitted. The X / Y axis translation and rotation angle are extracted as position offset parameters.

[0012] Optionally, the process of determining whether a preset overlap condition is met and generating position offset parameters and overlap status flags is as follows: The position offset parameters are compared with the preset overlap conditions: if the translation amount is ≤0.5μm and the rotation angle is ≤0.05°, a "overlapped" status flag is generated; otherwise, a "not overlapped" status flag is generated.

[0013] The present invention also provides an optical imaging system for implementing the imaging method of the optical imaging system described above, wherein the optical imaging mechanism includes: The imaging module includes a first camera and a second camera arranged side by side. The telescopic mechanism has its driving end fixedly connected to the imaging module, and moves the imaging module into or out of the bonding space via a lateral linear guide rail. An optical lens module includes a lens body, wherein the upper end surface of the lens body is provided with a first optical surface and the lower end surface is provided with a second optical surface, and both the first optical surface and the second optical surface are coated with a beam-splitting layer. The light source module emits a controllable power beam into the bonding space at a preset incident angle; The data processing unit is electrically connected to the first camera, the second camera, and the light source module, and performs image calibration, feature extraction, and offset calculation.

[0014] Optionally, the optical lens module further includes a piezoelectric ceramic layer, which is attached to the sidewall of the lens body and finely adjusts the pitch angle of the lens body based on real-time compensation parameters.

[0015] The present invention also provides a pick and place machine, including the optical imaging system described above, wherein the pick and place machine includes an upper top plate and a lower base arranged at an upper and lower interval, and a bonding space is formed between the upper top plate and the lower base; The upper top plate is provided with a first station for accommodating the bonding component, and the lower base is provided with a second station for accommodating the chip. The upper top plate and the lower base are respectively provided with positioning mechanisms for adjusting their positions.

[0016] Compared with existing technologies, this invention has the following advantages: The optical camera mechanism is laterally moved into the bonding space via a telescopic mechanism, initializing the dual cameras and positioning the top plate and bottom base to preset positions; based on synchronous reflection imaging of the dual optical surface beam-splitting layer: the first optical surface reflects the image of the bonding component to the first camera, and the second optical surface reflects the image of the chip to the second camera, generating a dual-target original image set; feature extraction is performed on the image set to identify key feature points of the chip and the bonding component and generate a coordinate dataset; the positional offset (including translation and rotation) of the feature coordinates of the two images is calculated to determine whether the overlap threshold is met; if they do not overlap, the driving mechanism adjusts the bonding space position and re-images; if they overlap, an alignment signal is output, forming a closed-loop control chain; this method achieves real-time dual-target capture through synchronous reflection of the beam-splitting layer, significantly improving alignment accuracy and greatly increasing chip bonding efficiency and yield. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] The structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0019] Figure 1 This is a schematic diagram of the optical path transmission of the optical camera system in this embodiment two; Figure 2 This is a schematic flowchart of the imaging method of the optical camera system in this embodiment. Detailed Implementation

[0020] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] In the description of this invention, it should be understood that the terms "upper," "lower," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. It should be noted that when a component is considered to be "connected" to another component, it can be directly connected to the other component or there may be a component positioned centrally in the connection.

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0023] Example 1: Combination Figure 2 As shown, this embodiment of the invention provides an imaging method for an optical camera system, comprising the following steps: S1, the optical camera mechanism is driven laterally through the telescopic mechanism to enter the bonding space between the upper top plate 3 and the lower base 4, and the camera and optical lens are initialized. At the same time, the upper top plate 3 and the lower base 4 are adjusted to the preset imaging position through the drive mechanism.

[0024] The optical camera mechanism is laterally moved into the bonding space by the telescopic mechanism to avoid vertical movement interference; the camera parameters (such as exposure time) and the optical lens beam splitting layer status (such as cleanliness calibration) are initialized; the upper top plate 3 and the lower base 4 are synchronously driven to the preset distance to ensure that the chip and the bonding component are within the effective depth of field of the optical lens.

[0025] S2, based on the beam-splitting layer characteristics of the optical lens, performs dual-target reflection imaging operation synchronously through the optical lens, enabling the camera to capture the reflected image of the chip and the reflected image of the bonding component, generating the original image dataset; utilizing the physical characteristics of the beam-splitting layer of the optical lens, the reflected light from the chip surface is reflected to the camera through the beam-splitting layer; the camera captures dual virtual images in a single exposure, generating the original image dataset, eliminating the timing error of step-by-step imaging, and improving image synchronization.

[0026] S3 performs preprocessing and feature extraction operations on the original image dataset, identifies key feature points of the chip and the bonding components, and generates a feature coordinate dataset.

[0027] Gray-level normalization and Gaussian filtering were performed on the original dataset to suppress ambient light noise; edge detection algorithms (such as the Canny operator) were used to extract the contour features of the chip pads and bonding component alignment marks; the pixel coordinates of the feature points were converted into millimeter-level physical coordinates with the bonding center as the origin, which is the algorithm model built into the camera system chip.

[0028] S4, based on the feature coordinate dataset, calculates the positional offset between two reflection images, determines whether the preset overlap conditions are met, and generates positional offset parameters and overlap status flags.

[0029] Based on the feature coordinate dataset, perform the following operations: Coordinate system alignment: Mapping the feature points of the chip and the bonding components to the same spatial coordinate system; Offset vector calculation: The least squares method is used to fit the translation (ΔX, ΔY) and rotation angle (Δθ) of the feature points; Overlap determination: If ΔX≤first preset value, ΔY≤second preset value and Δθ≤third preset value, then mark it as "overlapped".

[0030] S5, if the overlap status indicator indicates that they are not overlapped, then a drive command is generated according to the position offset parameter to control the drive mechanism to adjust the relative position of the upper top plate 3 and the lower base 4, and then return to execute to obtain the virtual image; if the overlap status indicator indicates that they are overlapped, then an alignment completion signal is output.

[0031] If the overlap status is "not overlapped", then: Generate pulse commands based on offset parameters (ΔX, ΔY, Δθ); The drive mechanism adjusts the position of the upper top plate 3 / lower base 4 (e.g., the X-axis moves by ΔX / 2, and the Y-axis moves by ΔY / 2). Trigger the optical system to re-execute imaging (return to S2).

[0032] If the signal indicates "already aligned," an electrical signal is output to lock the drive mechanism and initiate bonding. Technical effect: This forms a closed loop of "imaging-analysis-adjustment," iteratively approximating the aligned target.

[0033] The working principle of this invention is as follows: The optical camera mechanism is laterally moved into the bonding space via a telescopic mechanism, the dual cameras are initialized, and the upper top plate and lower base are positioned to preset positions. Based on synchronous reflection imaging using a dual-optical-surface beam-splitter layer: the first optical surface reflects the image of the bonding component to the first camera, and the second optical surface reflects the image of the chip to the second camera, generating a dual-target original image set. Feature extraction is performed on the image set to identify key feature points of the chip and the bonding component and generate a coordinate dataset. The positional offset (including translation and rotation) of the feature coordinates of the two images is calculated to determine whether the overlap threshold is met. If they do not overlap, the driving mechanism adjusts the position of the bonding space and re-images; if they overlap, an alignment signal is output, forming a closed-loop control chain. This method achieves real-time dual-target capture through synchronous reflection of the beam-splitter layer, significantly improving alignment accuracy and greatly improving chip bonding efficiency and yield. In this embodiment, step S1 specifically includes: S11, based on the height parameter of the bonding space, generates the drive path planning of the telescopic mechanism, drives the optical camera mechanism to move along the guide rail laterally through a linear motor, and collects displacement sensor data in real time to correct the position deviation until the central axis of the optical camera mechanism is aligned with the central plane of the bonding space.

[0034] The telescopic mechanism's movement path is planned based on the bonding space height; a linear motor drives the optical camera mechanism to move laterally along a linear guide rail; position deviations are fed back in real time through displacement sensors (such as grating rulers), and the trajectory is dynamically corrected using a PID control algorithm until the central axis of the camera mechanism coincides with the central plane of the bonding space. S12, start the camera and light source module of the optical camera mechanism, collect background light intensity data of the bonding space, dynamically adjust the output power of the light source based on the preset light intensity threshold, and generate anti-interference environment parameters.

[0035] Start the camera and LED light source module; collect the background light intensity of the bonding space through the photosensitive sensor; dynamically adjust the LED driving current according to the preset light intensity threshold (such as 500 lux for the target imaging area) so that the output light intensity and the ambient light are superimposed to meet the imaging requirements.

[0036] S13, load the beam splitter calibration coefficient of the optical lens to the data processing unit, calculate the real-time compensation parameters of the optical lens in combination with the anti-interference environment parameters, and control the piezoelectric fine adjuster to adjust the pitch angle of the optical lens.

[0037] Call the pre-stored beam splitter calibration coefficients (such as refractive index compensation values); combine the anti-interference environment parameters generated by S12 (such as actual light intensity distribution) to calculate the real-time pitch compensation angle of the optical lens; and use the piezoelectric fine-tuner to drive the lens support to fine-tune the angle.

[0038] S14: The encoder of the drive mechanism obtains the real-time distance data between the upper top plate 3 and the lower base 4, compares it with the target distance of the preset imaging position, calculates the position adjustment amount, and drives the upper top plate 3 and the lower base 4 to move towards each other to the target distance, and locks the position of the drive mechanism.

[0039] The real-time spacing data of the encoder of the drive mechanism is read and compared with the preset imaging target spacing to calculate the adjustment amount; the servo motor is controlled to drive the upper top plate 3 and the lower base 4 to move towards each other to the target spacing, and the self-locking device is activated to fix the position.

[0040] In this embodiment, the optical lens includes a lens body 21. The upper surface of the lens body 21 is provided with a first optical surface 22, and the lower surface of the lens body 21 is provided with a second optical surface 23. Both the first optical surface 22 and the second optical surface 23 are provided with a beam splitting layer. The first optical surface 22 and the second optical surface 23 are respectively set at a preset angle with the center line of the lens body 21.

[0041] It should be noted that the lens body 21 serves as the optical substrate, with its central axis aligned with the center plane of the bonding space; the first optical surface 22 is located on the upper end surface of the lens body 21, and the second optical surface 23 is located on the lower end surface, both of which are coated with a beam-splitting layer (such as a metal-dielectric composite film) to achieve beam splitting and reflection of the incident light; a preset angle is set (typically 45° tilt): The two optical surfaces are configured at a preset angle relative to the centerline of the lens body, forming an optical system with independently controllable reflection paths on both sides. The first and second optical surfaces, through specific angle designs, guide the reflected light from the connector and chip to different cameras, achieving synchronous imaging of the upper and lower targets. The beam-splitting layer preferably employs dielectric film coating technology, resulting in a high reflectivity effect when light is incident at 45±0.5°, while simultaneously ensuring a balanced distribution of transmittance and reflectivity. The preset angle setting establishes a fixed geometric relationship between the incident and reflected light paths, eliminating image interference through optical path geometric constraints, and ensuring the orthogonal projection characteristics of the reflected images from the chip and connector on the dual-camera imaging plane, providing a stable optical structural foundation for high-precision coordinate mapping.

[0042] In this embodiment, step S2 specifically includes the following steps: S21, based on the ambient light intensity sensor readings of the bonding space and the material reflectivity of the chip / bonding component, calculate the incident beam power range of the light source module, generate the beam intensity distribution matrix, and control the LED array to project polarized light with a wavelength of 520-650nm into the bonding space at an incident angle of 45° to form an anti-interference imaging light field.

[0043] An adaptive light source output strategy is dynamically generated by integrating ambient light intensity sensor detection and target material reflectivity calculation. First, based on real-time ambient light intensity data and the surface reflection characteristics of the chip / bond, the power boundary of the incident beam is precisely defined to ensure that the light source intensity avoids overexposure while effectively suppressing environmental interference. Further, polarized light technology (520-650nm wavelength) is employed with a 45° incident angle to construct a stable imaging light field: polarized light effectively reduces stray reflection interference, and within a specific wavelength range, it reduces material absorption loss while maintaining sensitivity in the visible light range. The generation of the beam intensity distribution matrix ensures that the light source energy distribution matches the spatial characteristics of the target area, and by actively adjusting the light source power and polarization state, the image signal-to-noise ratio is improved.

[0044] S22, the light reflected from the connector enters the first optical surface beam splitter at an incident angle θ1=45±0.5°, where the incident light is reflected by the beam splitter to generate a first reflected light that is transmitted to the first camera; simultaneously, the light reflected from the chip enters the second optical surface beam splitter at an incident angle θ2=45±0.5°, and its reflected light is transmitted to the second camera.

[0045] By constraining the incident angle of the beam-splitting layer to 45±0.5°, the reflected light from the bonding component and the chip is ensured to enter the beam-splitting layer with optimal efficiency. The beam-splitting layer design of the first and second optical surfaces precisely separates the reflected light paths: the reflected light from the bonding component is guided by the beam-splitting layer to the first camera, while the reflected light from the chip is simultaneously guided to the second camera. The incident angle is strictly controlled within the high reflectivity range of the beam-splitting layer (typically for dielectric film systems designed for 45°), maximizing reflection efficiency while avoiding energy loss or stray light crosstalk caused by deviations in the incident angle. The dual-path reflected light synchronous capture mechanism provides a synchronous image data foundation for subsequent dual-target position analysis.

[0046] S23, synchronously trigger the exposure sequence of the first camera and the second camera, and generate an initial image pair covering the chip pads and bonding assembly alignment marks by forming a positive interactive complementary viewpoint on the corresponding imaging plane through the first reflected light and the second reflected light.

[0047] By synchronously triggering dual-camera exposures, the imaging optical paths of the complementary viewpoints are strictly aligned in the temporal dimension. The first camera captures the top-down view generated by the reflected light from the bonding component, while the second camera captures the bottom-up view generated by the reflected light from the chip. The two viewpoints orthogonally cover the key areas of the chip pads and bonding component alignment marks. The complementary viewpoint design eliminates the field-of-view occlusion problem of single-camera imaging, ensuring full coverage of alignment feature points and avoiding the omission of feature points due to limited viewpoints. The multidimensional spatial information of the initial image pair provides the raw input for subsequent distortion correction and feature matching.

[0048] S24, call the real-time compensation parameters of the optical lens, and perform perspective transformation matrix operation on the initial image pair through the real-time compensation parameters to correct the geometric distortion caused by the fine adjustment of the lens angle and generate a pixel-level aligned calibration image pair.

[0049] Based on the real-time compensation parameters of the optical lenses, perspective transformation matrix operations are performed on the initial image to correct nonlinear geometric distortions caused by dynamic adjustments of the lens angles (such as pitch angle changes caused by piezoelectric fine-tuning). A mathematical model maps the physical deformation parameters of the lenses to pixel-level coordinate transformations of the image, eliminating trapezoidal distortion or imaging scaling effects introduced by lens angle deviations, and achieving pixel-level spatial alignment of the two camera images. This step ensures that subsequent feature point matching calculations are unaffected by the mechanical fine-tuning of the optical components, improving the accuracy of position offset calculations.

[0050] S25, the calibration image pairs are time-stamped and encoded with the real-time compensation parameters of the optical lens, the beam splitter calibration coefficient and the anti-interference environment parameters to generate an original image dataset including geometric distortion correction information, and stored in a cache queue.

[0051] The calibrated images, real-time compensation parameters, beam splitter calibration coefficients, and environmental anti-interference parameters are timestamped and encoded to construct a raw image dataset with complete contextual information. The timestamping synchronization mechanism ensures a one-to-one temporal correspondence between optical calibration data and image frames, avoiding parameter mismatch issues caused by data transmission delays or buffer misalignments.

[0052] In this embodiment, step S4 specifically includes the following steps: S41, map the chip feature coordinates and the bonding component feature coordinates in the feature coordinate dataset to a unified coordinate system with the center of the bonding space as the origin, and generate the chip standard coordinate set and the bonding component standard coordinate set.

[0053] S42 defines dynamic matching weights based on the structural similarity between the chip and the bonding component, and performs feature point pairing on the standard coordinate set of the chip and the standard coordinate set of the bonding component through a weighted nearest neighbor algorithm to generate a set of matching feature point pairs.

[0054] Weight definition: Assign a weight w=0.9 to feature points with a contour curvature >0.8 (such as pad corner points); assign a weight w=0.3 to flat edge points.

[0055] Matching execution: Weighted KD-tree nearest neighbor search is used to minimize the cost function: min∑wi•||Pi-Pj||2 generates a set of matching point pairs.

[0056] Technical function: to improve the matching priority of highly significant features and suppress noise interference.

[0057] S43, calculate the Euclidean distance and angular deviation of each pair of feature points in the matching feature point pair set, generate a local offset vector set, and remove abnormal point pairs that exceed the preset tolerance range.

[0058] For each pair of matching points: Euclidean distance: Angle deviation: S44 performs least-squares optimization on the local offset vector set, fits the global offset transformation matrix, and extracts the X / Y axis translation and rotation angle as position offset parameters.

[0059] Constructing a least squares optimization problem: The global translation (ΔX, ΔY) and rotation angle Δθ are obtained by solving.

[0060] S45, compare the position offset parameters with the preset overlap conditions: if the translation amount is ≤0.5μm and the rotation angle is ≤0.05°, then generate the "overlapped" status flag; otherwise, generate the "not overlapped" status flag.

[0061] Example 2: Combination Figure 1 As shown, the present invention also provides an optical imaging system for implementing the imaging method of the optical imaging system as described in Embodiment 1. The optical imaging mechanism includes: Imaging module 1 includes a first camera 11 and a second camera 12 arranged side by side.

[0062] The telescopic mechanism has its driving end fixedly connected to the imaging module 1, and moves the imaging module 1 into or out of the bonding space through the lateral linear guide rail.

[0063] The optical lens module 2 includes a lens body 21. The upper end surface of the lens body 21 is provided with a first optical surface 22 and the lower end surface is provided with a second optical surface 23. Both the first optical surface 22 and the second optical surface 23 are coated with a beam-splitting layer.

[0064] The light source module emits a controllable power beam into the bonding space at a 45° incident angle.

[0065] The data processing unit is electrically connected to the first camera 11, the second camera 12, and the light source module, and performs image calibration, feature extraction, and offset calculation.

[0066] The working principle of this invention is as follows: The optical imaging system drives the imaging module 1 into the bonding space laterally through the telescopic mechanism. The light source module projects a light beam to the bonding component and the chip at a 45° incident angle. The reflected light from the bonding component is split into a first reflected light ray by the first optical surface 22 beam splitting layer (the first camera 11 captures the image of the bonding component). The chip reflected light (the second reflected light ray) is processed by beam splitting on the second optical surface 23. The chip reflected light is reflected by the beam splitting layer to the second camera 12 to generate a chip image. The dual cameras are exposed synchronously to generate a pair of original images. The data processing unit performs optical path compensation, feature extraction, and offset calculation. If the calculated position offset exceeds the threshold, a driving command is generated to adjust the position of the bonding space and re-image until the offset meets the overlap condition, and then an alignment completion signal is output to form a closed-loop control.

[0067] In this embodiment, the optical lens module 2 also includes a piezoelectric ceramic layer, which is attached to the sidewall of the lens body 21 and finely adjusts the pitch angle of the lens body 21 based on real-time compensation parameters.

[0068] Example 3: Combination Figure 1 As shown, the present invention also provides a pick-and-place machine, including the optical imaging system of Embodiment 2 above. The pick-and-place machine includes an upper top plate 3 and a lower base 4 arranged at an upper and lower interval, with a bonding space formed between the upper top plate 3 and the lower base 4. The upper top plate 3 is provided with a first station for accommodating the bonding component, and the lower base 4 is provided with a second station for accommodating the chip. The upper top plate 3 and the lower base 4 are respectively provided with positioning mechanisms for adjusting their positions.

[0069] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An imaging method for an optical camera system, characterized in that, Includes the following steps: The optical camera mechanism is driven laterally through the telescopic mechanism into the bonding space between the upper top plate and the lower base, and the camera and optical lens are initialized. At the same time, the upper top plate and the lower base are adjusted to the preset imaging position through the drive mechanism. Based on the beam-splitting characteristics of optical lenses, dual-target reflection imaging is performed synchronously through optical lenses, enabling the camera to capture the reflection images of the chip and the bonding components, thereby generating the original image dataset. Preprocessing and feature extraction operations are performed on the original image dataset to identify key feature points of the chip and the bonding components, and to generate a feature coordinate dataset. Based on the feature coordinate dataset, the positional offset between the two reflection images is calculated, and it is determined whether the preset overlap condition is met, generating positional offset parameters and overlap status flags. If the overlap status flag indicates that the plates are not overlapped, a drive command is generated based on the position offset parameter to control the drive mechanism to adjust the relative position of the upper top plate and the lower base, and then the process returns to acquire the virtual image. If the overlap status flag indicates that the plates are overlapped, an alignment completion signal is output.

2. The imaging method of the optical camera system according to claim 1, characterized in that, The process of laterally driving the optical camera mechanism through a telescopic mechanism into the bonding space between the upper top plate and the lower base, initializing the camera and optical lenses, and simultaneously adjusting the upper top plate and the lower base to a preset imaging position through the drive mechanism specifically includes: The drive path planning of the telescopic mechanism is generated based on the height parameter of the bonding space. The optical camera mechanism is laterally driven by a linear motor to move along the guide rail, and displacement sensor data is collected in real time to correct the position deviation until the central axis of the optical camera mechanism is aligned with the central plane of the bonding space. The camera and light source modules of the optical camera mechanism are activated to collect background light intensity data of the bonding space. The output power of the light source is dynamically adjusted based on the preset light intensity threshold to generate anti-interference environmental parameters. The beam splitter calibration coefficient of the optical lens is loaded into the data processing unit, and the real-time compensation parameters of the optical lens are calculated in combination with the anti-interference environment parameters. The piezoelectric fine-tuner is then controlled to adjust the pitch angle of the optical lens. The encoder of the drive mechanism acquires real-time distance data between the upper top plate and the lower base, compares it with the target distance at the preset imaging position, calculates the position adjustment amount, and drives the upper top plate and the lower base to move towards each other to the target distance, and locks the position of the drive mechanism.

3. The imaging method of the optical camera system according to claim 2, characterized in that, The optical lens includes a lens body, a first optical surface is provided on the upper end surface of the lens body, and a second optical surface is provided on the lower end surface of the lens body. Both the first optical surface and the second optical surface are provided with a beam-splitting layer. The first optical surface and the second optical surface are respectively set at a preset angle to the center line of the lens body.

4. The imaging method of the optical camera system according to claim 3, characterized in that, The beam-splitting layer based on optical lenses synchronously performs dual-target reflection imaging operations through the optical lenses, enabling the camera to capture the reflected images of the chip and the bonding components, generating an original image dataset. Specifically, this includes the following steps: Based on the ambient light intensity sensor readings of the bonding space and the material reflectivity of the chip / bonding component, the incident beam power range of the light source module is calculated, a beam intensity distribution matrix is ​​generated, and the LED array is controlled to project polarized light with a wavelength of 520-650nm into the bonding space at an incident angle of 45° to form an anti-interference imaging light field. The incident light reflected by the connector enters the first optical surface beam splitter at an incident angle θ1=45±0.5°, where the incident light is reflected by the beam splitter to generate the first reflected light, which is transmitted to the first camera; simultaneously, the chip reflects light at an incident angle θ2=45±0.5° into the second optical surface beam splitter, where the reflected light is transmitted to the second camera. The exposure sequence of the first and second cameras is triggered synchronously, and the first and second reflected rays form a positive interactive complementary view on the corresponding imaging plane to generate an initial image pair covering the chip pads and bonding assembly alignment marks.

5. The imaging method of the optical camera system according to claim 4, characterized in that, The process involves forming a positive interactive complementary viewing angle on the corresponding imaging plane using the first and second reflected rays to generate a reflected image covering the chip pads and bonding assembly alignment marks. This process further includes: The real-time compensation parameters of the optical lens are called, and the perspective transformation matrix operation is performed on the initial image pair through the real-time compensation parameters to correct the geometric distortion caused by the fine adjustment of the lens angle and generate a pixel-level aligned calibration image pair. The calibration image pairs are time-stamped and encoded with the real-time compensation parameters of the optical lens, the beam splitter calibration coefficient, and the anti-interference environment parameters to generate an original image dataset including geometric distortion correction information, and stored in a cache queue.

6. The imaging method of the optical camera system according to claim 1, characterized in that, Based on the feature coordinate dataset, the positional offset between the reflected virtual image and the refracted virtual image is calculated, specifically including the following steps: The chip feature coordinates and the bonding component feature coordinates in the feature coordinate dataset are mapped to a unified coordinate system with the center of the bonding space as the origin, respectively, to generate the chip standard coordinate set and the bonding component standard coordinate set; Based on the structural similarity between the chip and the bonding component, a dynamic matching weight is defined. The weighted nearest neighbor algorithm is used to perform feature point pairing on the standard coordinate set of the chip and the standard coordinate set of the bonding component to generate a set of matching feature point pairs. Calculate the Euclidean distance and angular deviation of each pair of feature points in the matching feature point pair set, generate a local offset vector set, and remove outlier point pairs that exceed the preset tolerance range; The local offset vector set is optimized by least squares, and the global offset transformation matrix is ​​fitted. The X / Y axis translation and rotation angle are extracted as position offset parameters.

7. The imaging method of the optical camera system according to claim 1, characterized in that, The process of determining whether a preset overlap condition is met, generating position offset parameters and overlap status flags, is as follows: The position offset parameters are compared with the preset overlap conditions: if the translation amount is ≤0.5μm and the rotation angle is ≤0.05°, a "overlapped" status flag is generated; otherwise, a "not overlapped" status flag is generated.

8. An optical imaging system, characterized in that, An imaging method for implementing the optical camera system as described in any one of claims 1 to 7, wherein the optical camera mechanism comprises: The imaging module includes a first camera and a second camera arranged side by side. The telescopic mechanism has its driving end fixedly connected to the imaging module, and moves the imaging module into or out of the bonding space via a lateral linear guide rail. An optical lens module includes a lens body, wherein the upper end surface of the lens body is provided with a first optical surface and the lower end surface is provided with a second optical surface, and both the first optical surface and the second optical surface are coated with a beam-splitting layer. The light source module emits a controllable power beam into the bonding space at a preset incident angle; The data processing unit is electrically connected to the first camera, the second camera, and the light source module, and performs image calibration, feature extraction, and offset calculation.

9. The optical imaging system according to claim 8, characterized in that, The optical lens module also includes a piezoelectric ceramic layer, which is attached to the sidewall of the lens body and finely adjusts the pitch angle of the lens body based on real-time compensation parameters.

10. A pick-and-place machine, characterized in that, Including the optical imaging system as described in any one of claims 8-9, the pick-and-place machine includes an upper top plate and a lower base that are spaced apart vertically, and a bonding space is formed between the upper top plate and the lower base; The upper top plate is provided with a first station for accommodating the bonding component, and the lower base is provided with a second station for accommodating the chip. The upper top plate and the lower base are respectively provided with positioning mechanisms for adjusting their positions.

Images