A deep cavity chip eutectic bonding device and method

Abstract

The invention relates to the field of semiconductor packaging technology, and specifically discloses a deep cavity chip eutectic patch device and method, wherein the device comprises: a horizontal movement rough control module controls a suction nozzle mechanical arm to drive a to-be-patch component to move horizontally and obtain a monitoring image sequence below the to-be-patch component in real time; a horizontal movement precise control module accurately controls the suction nozzle mechanical arm to drive the to-be-patch component to move horizontally to a target horizontal coordinate based on the monitoring image sequence below the to-be-patch component and the patch position requirement; a vertical movement rough control module controls the suction nozzle mechanical arm to drive the to-be-patch component to move vertically downward and obtain a monitoring image sequence around the component; a vertical movement precise control module analyzes a component vertical movement stop timestamp based on the monitoring image sequence around the component, and controls the suction nozzle mechanical arm to drive the to-be-patch component to stop moving downward at the component vertical movement stop timestamp; the device is used to effectively improve the accuracy and quality of deep cavity chip eutectic patch and reduce patch errors.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor packaging, and in particular to a deep cavity chip eutectic bonding device and method. Background Art

[0002] In today's semiconductor device equipment manufacturing and integrated circuit industry, with the deep integration and rapid development of industrial Internet technology, higher and higher requirements are placed on the eutectic bonding process of deep cavity products. As a key link, the accuracy and efficiency of eutectic bonding of deep cavity products directly affect the quality and performance of the final product. The continuous miniaturization and integration of semiconductor device equipment have made the internal structure of the chip more and more complex, and the accuracy requirements for eutectic bonding have reached the micron level. At the same time, the large-scale production of integrated circuits requires efficient bonding technology to meet the huge market demand.

[0003] At present, in the field of eutectic bonding of deep cavity products, manual eutectic platforms still occupy a certain position. However, this traditional method has many obvious shortcomings. Manual operation relies on manual experience and skills, which not only makes it difficult to ensure the consistency and accuracy of each operation, but also often fails to meet the requirements of complex deep cavity structures and high precision. In addition, the manual eutectic platform lacks precise quantitative standards and real-time feedback mechanisms in horizontal and vertical movement control, making it difficult to achieve precise positioning. Moreover, due to the limited speed and efficiency of manual operation, it cannot meet the fast beat requirements of large-scale production. In the era of industrial Internet, intelligent and automated production modes have gradually become the mainstream. Manual eutectic platforms also have great difficulties in integrating with advanced production management systems and quality monitoring systems, and it is difficult to adapt to the trend of future industrial development.

[0004] Therefore, the present invention proposes a deep cavity chip eutectic bonding device and method. Summary of the invention

[0005] The present invention provides a deep cavity chip eutectic patch device and method, which can control the suction nozzle mechanical arm to drive the patch component to move horizontally through the horizontal movement rough control module, and obtain the monitoring image sequence below, which provides a basis for subsequent precise control. The horizontal movement is precisely controlled by the horizontal movement precise control module based on the monitoring image sequence and the patch position requirement, thereby improving the accuracy of the patch position in the horizontal direction. After the horizontal movement is completed, the vertical movement rough control module controls the vertical downward movement and obtains the monitoring image sequence around the component to prepare for precise stop movement. Finally, the vertical movement precise control module analyzes the stop timestamp based on the surrounding monitoring image sequence to achieve precise stop, thereby ensuring the patch accuracy in the vertical direction. The entire device combines the rough and precise control in the horizontal and vertical directions without relying on manual experience and skills, which not only ensures the consistency and accuracy of each operation, but also can cope with complex deep cavity structures and high-precision requirements, and realizes precise positioning of the patch position through precise quantitative standards and real-time feedback mechanisms, effectively improving the accuracy and quality of the deep cavity chip eutectic patch, reducing the patch error, and improving the production efficiency and product qualification rate.

[0006] The present invention provides a deep cavity chip eutectic bonding device, comprising:

[0007] The horizontal movement rough control module is used to control the nozzle robot arm to drive the component to be mounted to move horizontally, and at the same time, obtain the monitoring image sequence below in real time during the horizontal movement of the component to be mounted;

[0008] The horizontal movement precision control module is used to accurately control the nozzle robot arm to drive the component to be mounted to move horizontally to the target horizontal coordinate based on the monitoring image sequence below the component to be mounted and the mounting position requirements, and obtain the component horizontal movement control result;

[0009] The vertical movement rough control module is used to control the nozzle robot arm to drive the component to be mounted to move vertically downward when the horizontal movement control result of the component is obtained, and at the same time, obtain the monitoring image sequence around the component;

[0010] The vertical movement precision control module is used to analyze the component vertical movement stop timestamp based on the monitoring image sequence around the component, and control the suction nozzle robot arm to drive the component to be mounted to stop moving downward at the component vertical movement stop timestamp to obtain the component mounting result of the deep cavity chip.

[0011] Preferably, the horizontal movement rough control module comprises:

[0012] The horizontal movement rough control submodule is used to control the nozzle robot arm based on the position of the deep cavity chip cavity to drive the component to be mounted to move horizontally;

[0013] The lower monitoring image acquisition submodule is used to acquire the lower monitoring image sequence of the to-be-mounted component during horizontal movement in real time based on the camera set at the end of the component suction nozzle mechanical arm during the horizontal movement of the suction nozzle mechanical arm.

[0014] Preferably, the horizontal movement precision control module includes:

[0015] The cavity opening coordinate positioning submodule is used to determine the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system based on the monitoring image sequence below the component to be mounted;

[0016] The horizontal movement precision control submodule is used to control the nozzle robot arm in real time to drive the component to be mounted horizontally to the target horizontal coordinate based on the patch position requirements and the current cavity coordinates of the deep cavity chip cavity in the camera coordinate system, and obtain the component horizontal movement control result.

[0017] Preferably, the cavity orifice coordinate positioning submodule comprises:

[0018] An edge and contour extraction unit is used to perform edge detection on each lower monitoring image in the lower monitoring image sequence of the patch component to obtain an edge detection result of each lower monitoring image, extract the contour of the edge detection result based on the geometric characteristic deformation shape library of the cavity contour, and obtain all suspected cavity contours in each lower monitoring image;

[0019] An actual cavity orifice contour screening unit, used to screen out an actual cavity orifice contour sequence from all suspected cavity orifice contours in all lower monitoring images in the lower monitoring image sequence;

[0020] A coordinate conversion relationship establishing unit, used to establish a conversion relationship between an image coordinate system and a camera coordinate system;

[0021] The coordinate conversion unit is used to convert the cavity contour coordinates of the last cavity contour in the actual cavity contour sequence in the image coordinate system to the camera coordinate system based on the conversion relationship between the image coordinate system and the camera coordinate system, and obtain the current cavity coordinates of the deep cavity chip cavity in the camera coordinate system.

[0022] Preferably, the actual cavity outline screening unit comprises:

[0023] The contour arbitrary combination subunit is used to obtain a sequence of all suspected cavity orifice contours from all suspected cavity orifice contours in all lower monitoring images in the lower monitoring image sequence;

[0024] A feature point matching subunit is used to perform feature point matching on each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence to obtain all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence;

[0025] The contour sequence screening subunit is used to screen out the actual cavity orifice contour sequence from all the suspected cavity orifice contour sequences based on all the matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence.

[0026] Preferably, the profile sequence screening subunit comprises:

[0027] A deformation matrix construction end is used to construct a deformation matrix between each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence and the least square method;

[0028] A deformation matrix sequence generating end is used to sort the deformation matrices between all groups of adjacent suspected cavity orifice contours in the corresponding suspected cavity orifice contour sequence based on the order of all suspected cavity orifice contours in each suspected cavity orifice contour sequence to obtain a deformation matrix sequence for each suspected cavity orifice contour sequence;

[0029] The feature point spacing determination terminal is used to determine the distances between all different feature points of each suspected cavity orifice contour in each suspected cavity orifice contour sequence as the distances between all feature points of each suspected cavity orifice contour;

[0030] A longitudinal matching end is used to perform longitudinal matching on all feature point spacings of all suspected cavity orifice contours in the corresponding suspected cavity orifice contour sequence based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence, so as to obtain a sequence of all feature point spacings of the corresponding suspected cavity orifice contour sequence;

[0031] The contour sequence screening end is used to input all suspected cavity contour sequences and corresponding deformation matrix sequences and all feature point spacing sequences into the actual contour sequence screening model to obtain the actual cavity contour sequence.

[0032] Preferably, the horizontal movement precision control submodule includes:

[0033] The patch position projection positioning unit is used to determine the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system in real time based on the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system and the patch position requirements;

[0034] A camera-world coordinate conversion unit, used for converting the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system to obtain the target horizontal coordinates in the world coordinate system;

[0035] The horizontal movement precision control unit is used to control the nozzle robot arm in real time to drive the component to be mounted to move horizontally to the target horizontal coordinate, and obtain the component horizontal movement control result.

[0036] Preferably, the vertical movement rough control module comprises:

[0037] The vertical movement rough control submodule is used to control the nozzle robot arm to drive the component to be mounted to move vertically downward when the component horizontal movement control result is obtained;

[0038] The surrounding monitoring image acquisition submodule is used to acquire the surrounding monitoring image sequence of the component in real time during the vertical movement of the nozzle robot arm based on the camera set at the end of the component nozzle robot arm;

[0039] Each of the component surrounding monitoring images contained in the component surrounding monitoring image sequence contains a complete gap structure between the component to be mounted and the inner wall of the deep cavity chip cavity.

[0040] Preferably, the vertical movement precision control module includes:

[0041] The gap image segmentation and extraction submodule is used to extract the component-cavity gap image sequence from the monitoring image sequence around the component by using the gap image segmentation method based on morphology;

[0042] A movement stop timestamp determination submodule is used to analyze the component vertical movement stop timestamp based on the component-cavity gap image sequence;

[0043] The vertical movement stop control submodule is used to control the nozzle robot arm to drive the component to be mounted to stop moving downward at the component vertical movement stop timestamp to obtain the component mounting result of the deep cavity chip.

[0044] The present invention provides a deep cavity chip eutectic bonding method, which is applied to any of the above deep cavity chip eutectic bonding devices, comprising:

[0045] S1: Control the nozzle robot arm to drive the component to be mounted to move horizontally, and at the same time, obtain the monitoring image sequence below during the horizontal movement of the component to be mounted in real time;

[0046] S2: Based on the monitoring image sequence below the component to be mounted and the mounting position requirements, the nozzle robot arm is precisely controlled to drive the component to be mounted to move horizontally to the target horizontal coordinate to obtain the component horizontal movement control result;

[0047] S3: When the component horizontal movement control result is obtained, the nozzle robot arm is controlled to drive the component to be mounted to move vertically downward, and at the same time, a monitoring image sequence around the component is obtained;

[0048] S4: Analyze the timestamp of component vertical movement stop based on the monitoring image sequence around the component. At the timestamp of component vertical movement stop, control the nozzle robot arm to drive the component to be mounted to stop moving downward, and obtain the component mounting result of the deep cavity chip.

[0049] The beneficial effects of the present invention compared with the prior art are as follows: the horizontal movement rough control module can control the nozzle mechanical arm to drive the patch component to move horizontally, and obtain the monitoring image sequence below, which provides a basis for subsequent precise control. The horizontal movement precise control module can accurately control the horizontal movement based on the monitoring image sequence and the patch position requirements, thereby improving the accuracy of the patch position in the horizontal direction. After the horizontal movement is completed, the vertical movement rough control module controls the vertical downward movement and obtains the monitoring image sequence around the component to prepare for precise stop movement. Finally, the vertical movement precise control module analyzes the stop timestamp based on the surrounding monitoring image sequence to achieve precise stop, thereby ensuring the patch accuracy in the vertical direction. The entire device combines rough and precise control in the horizontal and vertical directions without relying on manual experience and skills, which not only ensures the consistency and accuracy of each operation, but also can cope with complex deep cavity structures and high-precision requirements, and realizes precise positioning of the patch position through precise quantitative standards and real-time feedback mechanisms, effectively improving the accuracy and quality of deep cavity chip eutectic patch, reducing patch errors, and improving production efficiency and product qualification rate.

[0050] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.

[0051] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0053] Figure 1 Schematic diagram of a deep cavity chip eutectic bonding device in an embodiment of the present invention;

[0054] Figure 2 4 is a flow chart of a deep cavity chip eutectic bonding method in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0056] Embodiment 1: The present invention provides a deep cavity chip eutectic bonding device, referring to Figure 1 ,include:

[0057] The horizontal movement rough control module is used to control the nozzle robot arm to drive the component to be mounted to move horizontally, and at the same time, obtain the monitoring image sequence below in real time during the horizontal movement of the component to be mounted;

[0058] The horizontal movement precision control module is used to accurately control the nozzle robot arm to drive the component to be mounted to move horizontally to the target horizontal coordinate based on the monitoring image sequence below the component to be mounted and the mounting position requirements, and obtain the component horizontal movement control result;

[0059] The vertical movement rough control module is used to control the nozzle robot arm to drive the component to be mounted to move vertically downward when the horizontal movement control result of the component is obtained, and at the same time, obtain the monitoring image sequence around the component;

[0060] The vertical movement precision control module is used to analyze the component vertical movement stop timestamp based on the monitoring image sequence around the component, and control the suction nozzle robot arm to drive the component to be mounted to stop moving downward at the component vertical movement stop timestamp to obtain the component mounting result of the deep cavity chip.

[0061] In this embodiment, the nozzle robot arm is a mechanical device that can grab and move the components to be mounted, and the end of the robot arm usually has a nozzle for adsorbing the components. For example, on an electronic component mounting production line, the robot arm is used to accurately move the chip for mounting.

[0062] In this embodiment, the component to be mounted refers to a part or component that needs to be installed or mounted to a specific position, for example, a heat sink that needs to be mounted to a deep cavity chip.

[0063] In this embodiment, the lower monitoring image sequence is a series of continuous images including the upper plane of the component to be mounted and the shape of the space below it, which are taken from the lower perspective during the horizontal movement of the nozzle robot arm. For example, during the mounting process, one lower image is taken per second to form a multi-frame image sequence.

[0064] In this embodiment, the patch position requirement is the requirement on the precise coordinates, angles, etc. of the patch component at the final installation position. For example, a chip is required to be accurately installed at a specific coordinate (x, y) on the circuit board and the rotation angle is 30 degrees.

[0065] In this embodiment, the target horizontal coordinate refers to the final accurate position coordinate to which the patch component needs to move in the horizontal direction, for example, the horizontal coordinate (100, 200).

[0066] In this embodiment, the component horizontal movement control result refers to the final state reached after the nozzle robot arm drives the component to be mounted to move horizontally. For example, the component is successfully moved to the specified horizontal position.

[0067] In this embodiment, the monitoring image sequence around the component is a series of continuous images captured from around the component when the nozzle robot arm drives the component to be mounted to move vertically downward. For example, multiple frames of images of the component are captured from the front, back, left, and right directions.

[0068] In this embodiment, the timestamp of component vertical movement stop refers to a specific time mark for determining that the nozzle robot arm should stop the vertical movement of the component to be mounted, for example, stopping the vertical movement at a time point accurate to milliseconds.

[0069] In this embodiment, the component mounting result of the deep cavity chip refers to the final installation state and effect of the component to be mounted in the deep cavity chip after the entire mounting operation is completed. For example, the chip is successfully and accurately mounted in the deep cavity chip, achieving good electrical connection and mechanical fixation.

[0070] The beneficial effects of the above technology are as follows: the horizontal movement rough control module can control the nozzle mechanical arm to drive the component to be patched to move horizontally, and obtain the monitoring image sequence below, which provides a basis for subsequent precise control. The horizontal movement precise control module can accurately control the horizontal movement based on the monitoring image sequence and the patch position requirements, thereby improving the accuracy of the patch position in the horizontal direction. After the horizontal movement is completed, the vertical movement rough control module controls the vertical downward movement and obtains the monitoring image sequence around the component to prepare for precise stop movement. Finally, the vertical movement precise control module analyzes the stop timestamp based on the surrounding monitoring image sequence to achieve precise stop, ensuring the patch accuracy in the vertical direction. The entire device combines rough and precise control in the horizontal and vertical directions without relying on manual experience and skills, which not only ensures the consistency and accuracy of each operation, but also can cope with complex deep cavity structures and high-precision requirements. The precise positioning of the patch position is achieved through precise quantitative standards and real-time feedback mechanisms, effectively improving the accuracy and quality of deep cavity chip eutectic patch, reducing patch errors, and improving production efficiency and product qualification rate.

[0071] Embodiment 2: Based on Embodiment 1, a horizontal movement rough control module includes:

[0072] The horizontal movement rough control submodule is used to control the nozzle robot arm to drive the component to be mounted to move horizontally based on the approximate orientation of the deep cavity chip cavity;

[0073] The lower monitoring image acquisition submodule is used to acquire the lower monitoring image sequence of the to-be-mounted component during horizontal movement in real time based on the camera set at the end of the component suction nozzle mechanical arm during the horizontal movement of the suction nozzle mechanical arm.

[0074] In this embodiment, the approximate position of the cavity opening of the deep cavity chip refers to a preliminary estimation or rough judgment of the direction of the cavity opening of the deep cavity chip, for example, roughly determining whether the cavity opening is located to the left, to the right, or in the center.

[0075] In this embodiment, the nozzle robot arm is controlled to drive the component to be mounted to move horizontally based on the approximate orientation of the cavity opening of the deep cavity chip, which means that the nozzle robot arm is directed to move the component to be mounted horizontally based on the understanding of the approximate direction of the cavity opening of the deep cavity chip. For example, if it is determined that the cavity opening is on the left, the robot arm is controlled to move horizontally with the component to the left.

[0076] In this embodiment, the end of the component suction nozzle robot arm refers to the end of the suction nozzle robot arm away from its control end and close to the suction nozzle. For example, it is the end where the suction nozzle is installed to grab the component.

[0077] The beneficial effects of the above technology are as follows: the horizontal movement rough control submodule is controlled based on the approximate orientation of the deep cavity chip cavity, providing preliminary direction guidance for horizontal movement and improving the efficiency of movement. The lower monitoring image acquisition submodule acquires the lower monitoring image sequence in real time by setting a camera at the end of the component suction nozzle robot arm, providing timely and accurate data support for subsequent precise control. This combination method can not only quickly achieve preliminary horizontal movement, but also make full preparations for subsequent precise adjustments, which helps to improve the efficiency and accuracy of the entire patch process, ensure patch quality, and reduce production costs.

[0078] Embodiment 3: Based on Embodiment 1, a horizontal movement precision control module includes:

[0079] The cavity opening coordinate positioning submodule is used to determine the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system based on the monitoring image sequence below the component to be mounted;

[0080] The horizontal movement precision control submodule is used to control the nozzle robot arm in real time to drive the component to be mounted horizontally to the target horizontal coordinate based on the patch position requirements and the current cavity coordinates of the deep cavity chip cavity in the camera coordinate system, and obtain the component horizontal movement control result.

[0081] In this embodiment, the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system refer to the coordinate values ​​of the position of the cavity opening of the deep cavity chip at this moment in the coordinate system established with the camera as the reference. For example, the origin of the camera coordinate system is set at the center of the camera lens, and the X, Y, and Z axes are determined according to certain rules. At this time, the position of the cavity opening of the deep cavity chip in this coordinate system can be represented by coordinates such as (X1, Y1, Z1).

[0082] The beneficial effects of the above technology are as follows: the cavity coordinate positioning submodule can determine the current cavity coordinates of the deep cavity chip cavity in the camera coordinate system based on the monitoring image sequence below, thereby realizing accurate perception of the cavity position. The horizontal movement precision control submodule controls the nozzle robot arm in real time to drive the component to be mounted to move horizontally to the target horizontal coordinate based on the patch position requirements and precise cavity coordinates, ensuring high precision of horizontal movement. This precise control method improves the matching degree between the component to be mounted and the cavity of the deep cavity chip in the horizontal direction, thereby effectively improving the accuracy and quality of the patch. It helps to reduce patch errors in the horizontal direction, improve production efficiency and product qualification rate, and reduce the defective rate and cost increase caused by inaccurate patch.

[0083] Embodiment 4: Based on Embodiment 3, the cavity opening coordinate positioning submodule includes:

[0084] An edge and contour extraction unit is used to perform edge detection on each lower monitoring image in the lower monitoring image sequence of the patch component to obtain an edge detection result of each lower monitoring image, extract the contour of the edge detection result based on the geometric characteristic deformation shape library of the cavity contour, and obtain all suspected cavity contours in each lower monitoring image;

[0085] An actual cavity orifice contour screening unit, used to screen out an actual cavity orifice contour sequence from all suspected cavity orifice contours in all lower monitoring images in the lower monitoring image sequence;

[0086] A coordinate conversion relationship establishing unit, used to establish a conversion relationship between an image coordinate system and a camera coordinate system;

[0087] The coordinate conversion unit is used to convert the cavity contour coordinates of the last cavity contour in the actual cavity contour sequence in the image coordinate system to the camera coordinate system based on the conversion relationship between the image coordinate system and the camera coordinate system, and obtain the current cavity coordinates of the deep cavity chip cavity in the camera coordinate system.

[0088] In this embodiment, edge detection is performed on each lower monitoring image in the lower monitoring image sequence of the component to be mounted to obtain the edge detection result of each lower monitoring image, which means that a specific algorithm is used to find the boundary between the object and the background of each image of the component to be mounted taken from below, and obtain information about these boundaries. For example, the contour lines of the object in the image are identified by the algorithm.

[0089] In this embodiment, the cavity opening contour is the outer shape of the opening of the deep cavity chip. For example, if the opening of the deep cavity chip is circular, the edge of the circle is the cavity opening contour.

[0090] In this embodiment, the geometric characteristic deformation shape library of the cavity contour is a pre-established collection of various possible cavity contour shapes and their changing forms in different situations. For example, the library contains shape models of circular and square cavities in different sizes and different deformation degrees.

[0091] In this embodiment, the contour of the edge detection result is extracted based on the geometric characteristics of the cavity contour, and all the suspected cavity contours in each lower monitoring image are obtained by deforming the shape library, which means that the part that may be the cavity contour is found from the result obtained by edge detection according to the information in the shape library. For example, parts similar to those in the shape library are screened out from the many detected contours, and they are considered to be suspected cavity contours.

[0092] In this embodiment, the lower monitoring image refers to an image taken from a lower perspective during the horizontal movement of the nozzle robot arm, including the upper plane of the component to be mounted and the shape of the space below it.

[0093] In this embodiment, the actual cavity opening contour sequence refers to an arrangement of a series of contours that are determined to truly belong to the cavity opening of the deep cavity chip after a series of screening and judgment, for example, a group of cavity opening contours that are confirmed to be arranged in chronological order after precise analysis.

[0094] In this embodiment, establishing the conversion relationship between the image coordinate system and the camera coordinate system refers to finding a mathematical rule so that the position of the pixel point in the image can correspond to the coordinate position in the three-dimensional space where the camera is located. For example, the (X, Y, Z) coordinates corresponding to a point (x, y) in the image in the camera coordinate system are obtained by calculation.

[0095] In this embodiment, based on the conversion relationship between the image coordinate system and the camera coordinate system, the cavity contour coordinates of the last cavity contour in the actual cavity contour sequence in the image coordinate system are converted to the camera coordinate system, and the current cavity coordinates of the cavity of the deep cavity chip in the camera coordinate system are obtained, that is, the cavity contour coordinates represented in the image are converted into the coordinates in the space where the camera is located by using the determined conversion rules, so as to determine the exact position of the cavity in the camera coordinate system. For example, the cavity contour coordinates (100, 200) represented in the image are converted to (50, 60, 70) in the camera coordinate system, given the known conversion relationship.

[0096] The beneficial effects of the above technology are as follows: the edge and contour extraction unit obtains all suspected cavity contours by performing edge detection and contour extraction on the monitoring image below, providing rich information for the subsequent screening of the actual cavity contour. The actual cavity contour screening unit can accurately screen out the actual cavity contour sequence from many suspected cavity contours, thereby improving the accuracy of cavity positioning. The conversion relationship between the image coordinate system and the camera coordinate system established by the coordinate conversion relationship establishment unit provides a key mathematical basis for coordinate conversion. The coordinate conversion unit converts the coordinates of the actual cavity contour in the image coordinate system to the camera coordinate system to obtain the accurate current cavity coordinates, providing accurate position data for precise control of horizontal movement. A series of operations of the entire cavity coordinate positioning submodule realizes the precise coordinate positioning of the cavity of the deep cavity chip, which helps to improve the effect of precise control of horizontal movement, thereby improving the accuracy and quality of the patch and ensuring the reliability and stability of the product.

[0097] Embodiment 5: Based on Embodiment 4, the actual cavity contour screening unit includes:

[0098] The contour arbitrary combination subunit is used to obtain all suspected cavity orifice contour sequences from all suspected cavity orifice contours in all lower monitoring images in the lower monitoring image sequence, wherein the suspected cavity orifice contour sequence includes any suspected cavity orifice contour in each lower monitoring image, and all suspected cavity orifice contour sequences cover all possible combinations of suspected cavity orifice contour sequences;

[0099] A feature point matching subunit is used to perform feature point matching on each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence to obtain all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence;

[0100] The contour sequence screening subunit is used to screen out the actual cavity orifice contour sequence from all the suspected cavity orifice contour sequences based on all the matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence.

[0101] In this embodiment, the suspected cavity opening contour sequence refers to a series of arrangements that are considered to be possible cavity opening contours of a deep cavity chip, for example, a set of cavity opening contours that have passed preliminary screening but have not yet been finally confirmed.

[0102] In this embodiment, feature point matching is performed on each group of adjacent suspected cavity orifice contours in each sequence of suspected cavity orifice contours, which means that representative feature points on two adjacent suspected cavity orifice contours are compared and matched in each sequence of suspected cavity orifice contours. For example, the center points of two adjacent suspected circular cavity orifice contours, specific points on the circumference, etc. are matched and compared.

[0103] In this embodiment, a matching feature point pair is a pair of two feature points that can correspond to each other when performing feature point matching. For example, a corner point on a suspected cavity orifice contour and a corner point at a corresponding position on an adjacent suspected cavity orifice contour constitute a matching feature point pair.

[0104] The beneficial effects of the above technology are as follows: the contour arbitrary combination subunit obtains all suspected cavity contour sequences, covering all possible combinations, and provides a comprehensive analysis object for subsequent screening. The feature point matching subunit performs feature point matching on adjacent suspected cavity contours to obtain matching feature point pairs, which provides a key basis for screening. The contour sequence screening subunit screens out the actual cavity contour sequence from many suspected cavity contour sequences based on matching feature point pairs, thereby improving the accuracy and reliability of screening. Through this series of operations, the actual cavity contour can be accurately screened out from many suspected cavity contours, providing a precise basis for subsequent coordinate positioning and precise control of horizontal movement, which helps to improve the accuracy and quality of the patch, and improve production efficiency and product qualification rate.

[0105] Embodiment 6: Based on Embodiment 5, the profile sequence screening subunit comprises:

[0106] A deformation matrix construction end is used to construct a deformation matrix between each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence and the least square method;

[0107] A deformation matrix sequence generating end is used to sort the deformation matrices between all groups of adjacent suspected cavity orifice contours in the corresponding suspected cavity orifice contour sequence based on the order of all suspected cavity orifice contours in each suspected cavity orifice contour sequence to obtain a deformation matrix sequence for each suspected cavity orifice contour sequence;

[0108] The feature point spacing determination terminal is used to determine the distances between all different feature points of each suspected cavity orifice contour in each suspected cavity orifice contour sequence as the distances between all feature points of each suspected cavity orifice contour;

[0109] A longitudinal matching end is used to perform longitudinal matching on all feature point spacings of all suspected cavity orifice contours in the corresponding suspected cavity orifice contour sequence based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence, so as to obtain a sequence of all feature point spacings of the corresponding suspected cavity orifice contour sequence;

[0110] The contour sequence screening end is used to input all suspected cavity contour sequences and corresponding deformation matrix sequences and all feature point spacing sequences into the actual contour sequence screening model to obtain the actual cavity contour sequence.

[0111] In this embodiment, based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence and the least squares method, a deformation matrix between each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence is constructed:

[0112] Obtain the coordinate information of all matching feature point pairs of each group of adjacent suspected cavity orifice contours. Then use the least squares method to calculate the mathematical model parameters that can best fit the changes of these feature point pairs through these coordinate data. These parameters ultimately constitute the deformation matrix that describes the shape changes between adjacent suspected cavity orifice contours. For example, assuming that the coordinates of the matching feature point pairs are (x1, y1) and (x2, y2), the least squares method is used to calculate the scaling ratio, rotation angle, and translation that can reflect the change from (x1, y1) to (x2, y2). These parameters constitute the elements of the deformation matrix. In this way, a deformation matrix between adjacent suspected cavity orifice contours is constructed to quantitatively describe the shape differences and changes between them.

[0113] In this embodiment, the deformation matrix sequence refers to an arrangement of a series of matrices describing shape changes between adjacent suspected cavity orifice contours. For example, these matrices can represent changes in size, angle, position, etc. of adjacent contours.

[0114] In this embodiment, the feature point spacing sequence includes the distance between two feature points belonging to the same suspected cavity orifice contour in all suspected cavity orifice contours, and the total number of feature point spacing sequences is the same as the total number of different feature points included in a single suspected cavity orifice contour.

[0115] In this embodiment, the actual contour sequence screening model is a mathematical model or algorithm for accurately screening out the true cavity contour sequence from a plurality of suspected cavity contour sequences. For example, by inputting various relevant data and features, the model can determine and output the sequence that best matches the actual cavity contour.

[0116] The training process of the actual contour sequence screening model usually requires the following types of training samples:

[0117] A large amount of image data of deep cavity chips in different scenarios, including various possible cavity shapes, sizes and positions;

[0118] For each image data, the true cavity orifice contour sequence and the corresponding suspected cavity orifice contour sequence are accurately marked, and the characteristic points, matching characteristic point pairs, deformation matrix, characteristic point spacing and other information are marked;

[0119] Deep cavity chip images under different lighting conditions, shooting angles and resolutions to increase the model's adaptability to diverse image environments;

[0120] Deep cavity chip images with various possible interference factors, such as impurities, shadows, reflections, etc., enable the model to accurately screen out the actual contour sequence in complex scenes;

[0121] Label samples with different accuracy requirements and error tolerances to train models to adapt to different practical application needs;

[0122] Image samples of deep cavity chips of different types, materials, and manufacturing processes to improve the versatility and generalization of the model;

[0123] By using these rich, diverse and accurately labeled training samples, the model can learn the characteristics and rules of the true cavity contour sequence in various situations, so that when faced with new unknown image data, it can accurately screen out the actual cavity contour sequence.

[0124] The beneficial effects of the above technology are as follows: the deformation matrix construction end constructs the deformation matrix by matching feature point pairs and the least squares method, which can quantify the deformation relationship between adjacent suspected cavity contours. The deformation matrix sequence generation end sorts the deformation matrix to form a deformation matrix sequence, which helps to fully understand the changing law of the contour. The feature point spacing determination end determines the feature point spacing of each suspected cavity contour, providing more detailed data for subsequent analysis. The longitudinal matching end performs longitudinal matching on the feature point spacing to obtain the feature point spacing sequence, further mining the features of the contour. The contour sequence screening end inputs a variety of data into the actual contour sequence screening model to accurately screen out the actual cavity contour sequence. The operation of the entire contour sequence screening subunit improves the accuracy and scientificity of the actual cavity contour screening, provides a reliable foundation for subsequent positioning and control, helps to improve the accuracy and quality of deep cavity chip eutectic bonding, and enhances the stability and reliability of the production process.

[0125] Embodiment 7: Based on Embodiment 3, the horizontal movement precision control submodule includes:

[0126] The patch position projection positioning unit is used to determine the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system in real time based on the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system and the patch position requirements;

[0127] A camera-world coordinate conversion unit, used for converting the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system to obtain the target horizontal coordinates in the world coordinate system;

[0128] The horizontal movement precision control unit is used to control the nozzle robot arm in real time to drive the component to be mounted to move horizontally to the target horizontal coordinate, and obtain the component horizontal movement control result.

[0129] In this embodiment, based on the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system and the patch position requirements, the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system are determined in real time, which means that according to the known current position of the cavity opening of the deep cavity chip in the camera coordinate system, combined with the specified patch position specific requirements, the coordinates of the vertical projection of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system are immediately calculated. For example, if the cavity opening coordinates are known to be (10, 20, 30), and the patch requirement is 5 units to the right of the cavity opening and 3 units in front, then the calculated projection coordinates on the horizontal plane may be (15, 23, 0).

[0130] In this embodiment, the newly obtained projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system are converted to the target horizontal coordinates in the world coordinate system, that is, the projection coordinates just calculated in the camera coordinate system are transformed into the horizontal coordinates in the world coordinate system through a specific conversion rule or formula. For example, the conversion relationship between the camera coordinate system and the world coordinate system is known, and (15,23,0) in the camera coordinate system is converted to (50,60,0) in the world coordinate system, and this (50,60,0) is the target horizontal coordinate.

[0131] The beneficial effects of the above technology are as follows: the patch position projection positioning unit can determine the projection coordinates according to the current cavity coordinates and patch position requirements, providing key intermediate data for subsequent coordinate conversion and precise control. The camera-world coordinate conversion unit converts the projection coordinates in the camera coordinate system to the target horizontal coordinates in the world coordinate system, realizing the conversion between different coordinate systems, making the control more in line with the actual scene requirements. The horizontal movement precision control unit controls the nozzle robot arm in real time to drive the component to be patched to move horizontally to the target horizontal coordinates, ensuring the high precision and accuracy of the horizontal movement. This step-by-step precise control method improves the matching accuracy of the component to be patched with the target patch position in the horizontal direction, and effectively improves the quality and accuracy of the patch. It helps to reduce the error of horizontal movement, improve production efficiency and product qualification rate, and reduce production costs and defective rates.

[0132] Embodiment 8: Based on Embodiment 1, the vertical movement rough control module includes:

[0133] The vertical movement rough control submodule is used to control the nozzle robot arm to drive the component to be mounted to move vertically downward when the component horizontal movement control result is obtained;

[0134] The surrounding monitoring image acquisition submodule is used to acquire the surrounding monitoring image sequence of the component in real time during the vertical movement of the nozzle robot arm based on the camera set at the end of the component nozzle robot arm;

[0135] Each of the component surrounding monitoring images contained in the component surrounding monitoring image sequence contains a complete gap structure between the component to be mounted and the inner wall of the deep cavity chip cavity.

[0136] The beneficial effects of the above technology are as follows: after the horizontal movement control is completed, the vertical movement rough control submodule promptly controls the nozzle robot arm to drive the component to be mounted to move vertically downward, ensuring the continuity of the mounting process. The surrounding monitoring image acquisition submodule uses the camera to obtain the surrounding monitoring image sequence of the component containing the complete gap structure in real time, providing rich visual information for subsequent precise stopping of movement. This method can achieve rapid response in the initial stage of vertical movement and make sufficient data preparation for subsequent precise control. It helps to improve the efficiency and accuracy of vertical movement, ensure the good match between the component to be mounted and the deep cavity chip cavity, and improve the mounting quality and production efficiency.

[0137] Embodiment 9: Based on Embodiment 1, a vertical movement precision control module includes:

[0138] The gap image segmentation and extraction submodule is used to extract the component-cavity gap image sequence from the monitoring image sequence around the component by using the gap image segmentation method based on morphology;

[0139] A movement stop timestamp determination submodule is used to analyze the component vertical movement stop timestamp based on the component-cavity gap image sequence;

[0140] The vertical movement stop control submodule is used to control the nozzle robot arm to drive the component to be mounted to stop moving downward at the component vertical movement stop timestamp to obtain the component mounting result of the deep cavity chip.

[0141] In this embodiment, the morphological-based gap image segmentation method is a technique that uses mathematical morphology principles to process and segment an image. It extracts the gap image region in the image through a series of morphological operations, such as dilation, erosion, opening operation, closing operation, etc. For example, in a binary image, the noise is removed by the erosion operation, and then the general shape of the target is restored by the dilation operation.

[0142] In this embodiment, extracting a component-cavity gap image sequence from a monitoring image sequence around a component using a morphological gap image segmentation method means using the above-mentioned image segmentation method to separate images showing the gap between the component and the cavity from a series of monitoring images taken around the component, and arranging these images in order into a sequence. For example, a series of images are processed to retain only images showing the gap portion, forming a component-cavity gap image sequence.

[0143] In this embodiment, the component-cavity gap image sequence is an ordered set of images specifically showing the gap between the component and the cavity, for example, multiple images of the gap around the component taken in chronological order.

[0144] In this embodiment, the component vertical movement stop timestamp is analyzed based on the component-cavity gap image sequence, which means that the specific time point at which the nozzle robot arm should stop driving the component vertical movement is determined by analyzing and calculating the gap image sequence. For example, according to the trend of the gap gradually getting smaller, when the gap reaches a certain minimum value or meets a specific condition, the corresponding time is the stop timestamp.

[0145] The beneficial effects of the above technology are as follows: the gap image segmentation and extraction submodule uses a morphology-based gap image segmentation method to accurately extract the component-cavity gap image sequence, providing an accurate data basis for subsequent analysis. The movement stop timestamp determination submodule can accurately analyze the component vertical movement stop timestamp based on the component-cavity gap image sequence, ensuring that the timing of stopping the movement is just right. The vertical movement stop control submodule controls the nozzle robot arm to stop moving at the accurate timestamp, thereby obtaining the ideal deep cavity chip component patch result. Through this series of operations, precise control of vertical movement is achieved, effectively avoiding excessive or insufficient movement, and improving the accuracy and quality of the patch. It helps to reduce patch errors, improve the product qualification rate and performance stability, while improving production efficiency and reducing production costs.

[0146] Embodiment 10: The present invention provides a deep cavity chip eutectic bonding method, which is applied to any one of the deep cavity chip eutectic bonding devices in Embodiments 1 to 9, with reference to Figure 2 ,include:

[0147] S1: Control the nozzle robot arm to drive the component to be mounted to move horizontally, and at the same time, obtain the monitoring image sequence below during the horizontal movement of the component to be mounted in real time;

[0148] S2: Based on the monitoring image sequence below the component to be mounted and the mounting position requirements, the nozzle robot arm is precisely controlled to drive the component to be mounted to move horizontally to the target horizontal coordinate to obtain the component horizontal movement control result;

[0149] S3: When the component horizontal movement control result is obtained, the nozzle robot arm is controlled to drive the component to be mounted to move vertically downward, and at the same time, a monitoring image sequence around the component is obtained;

[0150] S4: Analyze the timestamp of component vertical movement stop based on the monitoring image sequence around the component. At the timestamp of component vertical movement stop, control the nozzle robot arm to drive the component to be mounted to stop moving downward, and obtain the component mounting result of the deep cavity chip.

[0151] The beneficial effects of the above technology are as follows: through step S1, the nozzle mechanical arm can be controlled to drive the component to be patched to move horizontally, and the monitoring image sequence below is obtained, which provides a basis for subsequent precise control. And through step S2, the horizontal movement is precisely controlled based on the monitoring image sequence and the patch position requirements, which improves the accuracy of the patch position in the horizontal direction. Then, through step S3, after the horizontal movement is completed, the vertical downward movement is controlled and the monitoring image sequence around the component is obtained to prepare for the precise stop movement. Finally, through step S4, the stop timestamp is analyzed based on the surrounding monitoring image sequence to achieve precise stop, ensuring the patch accuracy in the vertical direction. The entire device combines rough and precise control in the horizontal and vertical directions without relying on manual experience and skills, which not only ensures the consistency and accuracy of each operation, but also can cope with complex deep cavity structures and high-precision requirements. The precise positioning of the patch position is achieved through precise quantitative standards and real-time feedback mechanisms, effectively improving the accuracy and quality of deep cavity chip eutectic patch, reducing patch errors, and improving production efficiency and product qualification rate.

[0152] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A deep cavity chip eutectic bonding device, characterized in that: include: The horizontal movement rough control module is used to control the nozzle robot arm to drive the component to be mounted to move horizontally, and at the same time, obtain the monitoring image sequence below in real time during the horizontal movement of the component to be mounted; Horizontal movement precision control module, including: The cavity opening coordinate positioning submodule is used to determine the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system based on the monitoring image sequence below the component to be mounted; The horizontal movement precision control submodule is used to control the nozzle robot arm in real time to drive the component to be mounted to move horizontally to the target horizontal coordinate based on the patch position requirements and the current cavity opening coordinates of the deep cavity chip in the camera coordinate system, and obtain the component horizontal movement control result; A vertical movement rough control module is used to control the nozzle robot arm to drive the component to be mounted to move vertically downward when the component horizontal movement control result is obtained, and at the same time, obtain a monitoring image sequence around the component, wherein each monitoring image around the component contained in the monitoring image sequence around the component contains a complete gap structure between the component to be mounted and the inner wall of the deep cavity chip cavity; Vertical movement precision control module, including: The gap image segmentation and extraction submodule is used to extract the gap image sequence of the component to be mounted - the inner wall of the cavity from the monitoring image sequence around the component by using the gap image segmentation method based on morphology; A movement stop timestamp determination submodule is used to analyze the vertical movement stop timestamp of the component based on the image sequence of the gap between the component to be mounted and the inner wall of the cavity; The vertical movement stop control submodule is used to control the nozzle robot arm to drive the component to be mounted to stop moving downward at the component vertical movement stop timestamp to obtain the component mounting result of the deep cavity chip.

2. The deep cavity chip eutectic bonding device according to claim 1, characterized in that: Horizontal movement coarse control module, including: The horizontal movement rough control submodule is used to control the nozzle robot arm based on the position of the deep cavity chip cavity to drive the component to be mounted to move horizontally; The lower monitoring image acquisition submodule is used to acquire the lower monitoring image sequence of the to-be-mounted component during horizontal movement in real time based on the camera set at the end of the component suction nozzle mechanical arm during the horizontal movement of the suction nozzle mechanical arm.

3. The deep cavity chip eutectic bonding device according to claim 1, characterized in that: The cavity coordinate positioning submodule includes: An edge and contour extraction unit is used to perform edge detection on each lower monitoring image in the lower monitoring image sequence of the patch component to obtain an edge detection result of each lower monitoring image, extract the contour of the edge detection result based on the geometric characteristic deformation shape library of the cavity contour, and obtain all suspected cavity contours in each lower monitoring image; An actual cavity orifice contour screening unit, used to screen out an actual cavity orifice contour sequence from all suspected cavity orifice contours in all lower monitoring images in the lower monitoring image sequence; A coordinate conversion relationship establishing unit, used to establish a conversion relationship between an image coordinate system and a camera coordinate system; The coordinate conversion unit is used to convert the cavity contour coordinates of the last cavity contour in the actual cavity contour sequence in the image coordinate system to the camera coordinate system based on the conversion relationship between the image coordinate system and the camera coordinate system, and obtain the current cavity coordinates of the deep cavity chip cavity in the camera coordinate system.

4. The deep cavity chip eutectic bonding device according to claim 3, characterized in that: Actual cavity contour screening unit, including: The contour arbitrary combination subunit is used to obtain a sequence of all suspected cavity orifice contours from all suspected cavity orifice contours in all lower monitoring images in the lower monitoring image sequence; A feature point matching subunit is used to perform feature point matching on each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence to obtain all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence; The contour sequence screening subunit is used to screen out the actual cavity orifice contour sequence from all the suspected cavity orifice contour sequences based on all the matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence.

5. The deep cavity chip eutectic bonding device according to claim 4, characterized in that: Profile sequence screening subunit, including: A deformation matrix construction end is used to construct a deformation matrix between each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence and the least square method; A deformation matrix sequence generating end is used to sort the deformation matrices between all groups of adjacent suspected cavity orifice contours in the corresponding suspected cavity orifice contour sequence based on the order of all suspected cavity orifice contours in each suspected cavity orifice contour sequence to obtain a deformation matrix sequence for each suspected cavity orifice contour sequence; The feature point spacing determination terminal is used to determine the distances between all different feature points of each suspected cavity orifice contour in each suspected cavity orifice contour sequence as the distances between all feature points of each suspected cavity orifice contour; A longitudinal matching end is used to perform longitudinal matching on all feature point spacings of all suspected cavity orifice contours in the corresponding suspected cavity orifice contour sequence based on all matching feature point pairs of each group of adjacent suspected cavity orifice contours in each suspected cavity orifice contour sequence, so as to obtain a sequence of all feature point spacings of the corresponding suspected cavity orifice contour sequence; The contour sequence screening end is used to input all suspected cavity contour sequences and corresponding deformation matrix sequences and all feature point spacing sequences into the actual contour sequence screening model to obtain the actual cavity contour sequence.

6. The deep cavity chip eutectic bonding device according to claim 1, characterized in that: Horizontal movement precision control submodule, including: The patch position projection positioning unit is used to determine the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system in real time based on the current cavity opening coordinates of the deep cavity chip cavity opening in the camera coordinate system and the patch position requirements; A camera-world coordinate conversion unit, used for converting the projection coordinates of the target patch position of the component to be patched on the horizontal plane of the camera coordinate system to obtain the target horizontal coordinates in the world coordinate system; The horizontal movement precision control unit is used to control the nozzle robot arm in real time to drive the component to be mounted to move horizontally to the target horizontal coordinate, and obtain the component horizontal movement control result.

7. The deep cavity chip eutectic bonding device according to claim 1, characterized in that: Vertical movement coarse control module, including: The vertical movement rough control submodule is used to control the nozzle robot arm to drive the component to be mounted to move vertically downward when the component horizontal movement control result is obtained; The surrounding monitoring image acquisition submodule is used to acquire the surrounding monitoring image sequence of the component in real time during the vertical movement of the nozzle robot arm based on the camera set at the end of the component nozzle robot arm.

8. A deep cavity chip eutectic bonding method of the deep cavity chip eutectic bonding device according to any one of claims 1 to 7, comprising: S1: Control the nozzle robot arm to drive the component to be mounted to move horizontally, and at the same time, obtain the monitoring image sequence below during the horizontal movement of the component to be mounted in real time; S2: Based on the monitoring image sequence below the component to be mounted and the mounting position requirements, the nozzle robot arm is precisely controlled to drive the component to be mounted to move horizontally to the target horizontal coordinate to obtain the component horizontal movement control result; S3: When the component horizontal movement control result is obtained, the nozzle robot arm is controlled to drive the component to be mounted to move vertically downward, and at the same time, a monitoring image sequence around the component is obtained; S4: Analyze the component vertical movement stop timestamp based on the monitoring image sequence around the component, and control the nozzle robot arm to drive the component to be mounted to stop moving downward at the component vertical movement stop timestamp to obtain the component mounting result of the deep cavity chip.

Images