A Visual Intelligent Adjustment Method and System for a Full-Automatic Chip Mounter

Abstract

The present application discloses a visual intelligent adjustment method and system for a full-automatic chip mounter, which relates to the field of semiconductor back-end packaging manufacturing. The method includes: based on the camera in the full-automatic chip mounter, collecting images of features to be recognized in real time to obtain real-time images; based on multiple light sources and real-time images in the full-automatic chip mounter, determining the optimal light source and optimal brightness of the features to be recognized; obtaining the feature images to be recognized under the optimal light source and optimal brightness to obtain optimal images, and determining the template regions of the optimal images; based on the template regions of the optimal images, determining the regions of interest of the optimal images; based on the regions of interest of the optimal images, cropping the optimal images to obtain the region-of-interest images of the features to be recognized; and adjusting the camera based on the region-of-interest images of the features to be recognized. The present application accurately obtains the position and angle of the features to be recognized through the region-of-interest images, improves the operation accuracy of the features to be recognized, and improves the quality of chip mounting.

Description

Technical Field

[0001] This application relates to the technical field of semiconductor back-end packaging manufacturing, and particularly to a method and system for visual intelligent adjustment of a full-automatic die bonder. Background Art

[0002] During the die loading process of existing die bonders for semiconductor back-end packaging, manual adjustment of lighting, manual template making, and manual zooming are often relied on. There are problems such as complex operation, low precision, low efficiency, difficulty in judging quality, and some cannot save and call recipe management. At the same time, due to the errors of manual operation, the accuracy of the loading position and angle may be unstable, affecting the quality and performance of the die bonder for semiconductor equipment. Therefore, how to achieve visual intelligence of the die bonder, improve the stability and quality of the loading position and angle accuracy, improve the efficiency during visual replacement of different varieties, and reduce the errors of manual operation are urgent problems to be solved in the current field of back-end packaging equipment. Summary of the Invention

[0003] The purpose of this application is to provide a method and system for visual intelligent adjustment of a full-automatic die bonder, which can accurately obtain the position and angle of the feature to be recognized according to a clear image of the feature to be recognized, improve the operation accuracy of the feature to be recognized, and improve the quality of die loading.

[0004] To achieve the above purpose, this application provides the following solutions:

[0005] In the first aspect, this application provides a method for visual intelligent adjustment of a full-automatic die bonder, including:

[0006] Based on a camera in the full-automatic die bonder, images of the feature to be recognized are collected in real time to obtain real-time images; the feature to be recognized is a wafer or a substrate.

[0007] Based on multiple light sources in the full-automatic die bonder and the real-time images, the optimal light source and optimal brightness of the feature to be recognized are determined.

[0008] An image of the feature to be recognized under the optimal light source and optimal brightness is obtained to get an optimal image, and the template area of the optimal image is determined.

[0009] Based on the template area of the optimal image, the region of interest of the optimal image is determined.

[0010] Based on the region of interest of the optimal image, the optimal image is cropped to obtain an image of the region of interest of the feature to be recognized; and the camera is adjusted based on the image of the region of interest of the feature to be recognized.

[0011] In the second aspect, this application provides a system for visual intelligent adjustment of a full-automatic die bonder, including:

[0012] An image acquisition module, which is used to collect images of features to be recognized in real time based on a camera in a fully automatic chip mounter, so as to obtain real-time images; the features to be recognized are wafers or substrates.

[0013] An optimal light source and optimal brightness determination module, which is connected to the image acquisition module, and is used to determine the optimal light source and optimal brightness of the features to be recognized based on multiple light sources in the fully automatic chip mounter and the real-time images.

[0014] A template area determination module, which is connected to the optimal light source and optimal brightness determination module, and is used to obtain images of features to be recognized under the optimal light source and optimal brightness, so as to obtain optimal images, and determine the template area of the optimal images.

[0015] An interested area determination module, which is connected to the template area determination module, and is used to determine the interested area of the optimal image based on the template area of the optimal image.

[0016] An adjustment module, which is connected to the interested area determination module, and is used to crop the optimal image based on the interested area of the optimal image to obtain an interested image of the features to be recognized; and adjust the camera based on the interested image of the features to be recognized.

[0017] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0018] The present application provides a method and system for visual intelligent adjustment of a fully automatic chip mounter. Based on a camera in the fully automatic chip mounter, images of wafers or substrates in the fully automatic chip mounter are collected in real time. When determining the optimal light source and optimal brightness, quick feedback is provided, improving work efficiency. At the same time, the cumbersome manual adjustment is solved, and the manual adjustment error is avoided. The optimal image is obtained and the template area of the optimal image is determined, avoiding the difference in the template center position when different operators select the template area; by determining the interested area, the optimal field of view range is achieved. By obtaining the interested image to adjust the camera, the images of the wafers or substrates collected are clearer and more accurate. The present application accurately obtains the position and angle of the features to be recognized through the interested image, improves the subsequent operation accuracy of the features to be recognized, and improves the quality of chip mounting. Description of the Drawings

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0020] Figure 1Schematic diagram of the structure of a fully automatic chip mounter in an embodiment of the present application.

[0021] Figure 2 Schematic diagram of the structure of the camera part in the fully automatic chip mounter in an embodiment of the present application.

[0022] Figure 3 Application environment diagram of a vision intelligent adjustment method for a fully automatic chip mounter in an embodiment of the present application.

[0023] Figure 4 Flow schematic diagram of a vision intelligent adjustment method for a fully automatic chip mounter provided in an embodiment of the present application.

[0024] Figure 5 Functional module schematic diagram of a vision intelligent adjustment system for a fully automatic chip mounter provided in an embodiment of the present application.

[0025] Figure 6 Schematic diagram of the structure of a computer device provided in an embodiment of the present application.

[0026] Reference numerals:

[0027] Feeding part - 1, Track conveying part - 2, Gluing part - 3, Gluing camera part - 4, Bonding part - 5, Bonding camera part - 6, Wafer camera part - 7, Wafer platform - 8, Thimble part - 9, Frame part - 10, Discharging part - 11, High - pixel camera - 12, Large - field high - precision lens barrel - 13, Supporting multiple light sources - 14. Detailed implementation manners

[0028] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0029] To make the objectives, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0030] As Figure 1 shown, the fully automatic chip mounter includes a feeding part 1, a track conveying part 2, a gluing part 3, a gluing camera part 4, a bonding part 5, a bonding camera part 6, a wafer camera part 7, a wafer platform 8, a thimble part 9, a frame part 10, a discharging part 11, and a gluing cushion block.

[0031] Among them, the loading section 1 feeds the substrate into the production line. The track conveying section 2 conveys the substrate through the gripper within the track. The dispensing section 3 applies glue to the substrate. The dispensing camera section 4 locates the specific position of a single pad on the substrate when the substrate is conveyed to the camera section. The dispensing pad supports the frame when dispensing glue or drawing glue on the substrate. The bonding section 5 bonds the wafer to the frame. The bonding camera section 6 locates the specific position of the pads on the substrate when the substrate is conveyed to the bonding section. The wafer camera section 7 locates the specific position of the chips on the wafer. The wafer platform 8 moves the wafer in the X, Y, and R directions. The ejector pin section 9 ejects the chips from the wafer so that the bonding head can pick up the chips. The unloading section 11 removes the packaged devices from the production line.

[0032] The main working principle of the full-automatic die bonder is as follows: After sucking the wafer located by the wafer camera section 7 from the wafer platform 8, it is mounted on the lead frame or substrate that has been located by the bonding camera section 6 and has been dispensed with glue (or formed by dispensing tin, or dipped in flux). By obtaining clear and accurate images of the wafer and the substrate in this application, the mounting is more precise, improving the die bonding quality.

[0033] The dispensing camera section 4, the bonding camera section 6, and the wafer camera section 7 of the full-automatic die bonder adopt the visual intelligent adjustment method of the full-automatic die bonder to realize the intelligence of the vision of the full-automatic die bonder. The dispensing camera section 4, the bonding camera section 6, and the wafer camera section 7 mainly consist of Figure 2 a high-pixel camera 12, a large-field-of-view and high-precision lens barrel 13, and a variety of supporting light sources 14 as shown.

[0034] The visual intelligent adjustment method of the full-automatic die bonder provided by the embodiments of this application can be applied to Figure 3In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set separately, integrated on the server 104, placed on the cloud or other servers. The terminal 102 can send the image of the feature to be recognized to the server 104. After receiving the image of the feature to be recognized, for the image of the feature to be recognized, the server 104 obtains a real-time image, and based on multiple light sources in the fully automatic chip mounter and the real-time image, determines the optimal light source and optimal brightness of the feature to be recognized; obtains the image of the feature to be recognized under the optimal light source and optimal brightness to obtain an optimal image, and determines the template area of the optimal image; based on the template area of the optimal image, determines the region of interest of the optimal image; based on the region of interest of the optimal image, crops the optimal image to obtain the region of interest image of the feature to be recognized; and adjusts the camera based on the region of interest image of the feature to be recognized. The server 104 can feedback the image of the feature to be recognized collected after adjustment to the terminal 102. In addition, in some embodiments, the visual intelligent adjustment method of the fully automatic chip mounter can also be implemented independently by the server 104 or the terminal 102.

[0035] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0036] In an exemplary embodiment, as Figure 4 shown, a visual intelligent adjustment method for a fully automatic chip mounter is provided. This method is executed by a computer device, and can be specifically executed independently by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiments of the present application, taking this method applied to Figure 1 the server 104 in it as an example for illustration, it includes the following steps 201 to step 205. Among them:

[0037] Step 201, based on the camera in the fully automatic chip mounter, collect the image of the feature to be recognized in real time to obtain a real-time image; the feature to be recognized is a wafer or a substrate.

[0038] Step 202, based on multiple light sources in the fully automatic chip mounter and the real-time image, determine the optimal light source and optimal brightness of the feature to be recognized.

[0039] Step 203: Obtain the feature image to be recognized under the optimal light source and optimal brightness, obtain the optimal image, and determine the template area of the optimal image.

[0040] Step 204: Determine the region of interest of the optimal image based on the template area of the optimal image.

[0041] Step 205: Crop the optimal image based on the region of interest of the optimal image to obtain the region-of-interest image of the feature to be recognized; and adjust the camera based on the region-of-interest image of the feature to be recognized.

[0042] In another exemplary embodiment of the present application, the above step 202 specifically includes: Based on multiple light sources and real-time images in a fully automatic chip mounter, using the golden section search algorithm and the band-pass algorithm, determine the optimal light source and optimal brightness of the feature to be recognized. Enable the fully automatic chip mounter to automatically adjust parameters such as the brightness and color temperature of the light according to the characteristics of the wafer and lead frame, such as material and color, to ensure that the vision system of the chip mounter can accurately capture the image information of the wafer and the pins or base islands of the lead frame. Specifically, it includes the following steps 301 to 306.

[0043] Step 301: For the nth iteration of any light source, based on the nth brightness interval of the light source, use the golden section search algorithm to determine the first brightness in the nth brightness interval and the second brightness in the nth brightness interval. n > 0. In the present application, the first brightness interval is [0, 255], and the golden section ratio is usually 0.618. Determine the first brightness and the second brightness in the first brightness interval through the golden section ratio.

[0044] Step 302: Obtain the feature image to be recognized at the first brightness in the nth brightness interval and the feature image to be recognized at the second brightness in the nth brightness interval.

[0045] Step 303: Based on the feature image to be recognized at the first brightness in the nth brightness interval and the feature image to be recognized at the second brightness in the nth brightness interval, use the band-pass algorithm to determine the clarity of the feature image to be recognized at the first brightness in the nth brightness interval and the clarity of the feature image to be recognized at the second brightness in the nth brightness interval.

[0046] Step 304: Based on the clarity, determine the (n + 1)th brightness interval according to the nth brightness interval, the first brightness in the nth brightness interval, and the second brightness in the nth brightness interval.

[0047] Step 305: Determine whether n is less than the total number of iterations. If so, perform the (n + 1)th iteration; if not, determine the clarity maximum value of the light source based on the clarity of the feature image to be recognized at the first brightness from the first brightness interval to the nth brightness interval and the clarity of the feature image to be recognized at the second brightness.

[0048] Step 306: Determine the optimal light source and optimal brightness of the feature to be recognized based on the maximum clarity values of multiple light sources.

[0049] In another exemplary embodiment of the present application, there are several metrics for evaluating image clarity in current image processing algorithms as follows:

[0050] I. Gradient magnitude: The gradient magnitude represents the rate of change of pixel values in an image. A clear image usually has more edges and details, so its gradient magnitude is larger. The clarity of an image can be evaluated by calculating the gradient of the image, and a larger gradient magnitude usually indicates a clearer image.

[0051] II. Energy spectrum: The energy spectrum represents the energy distribution of different spatial frequencies in an image. A clear image usually has more high-frequency components, so its energy spectrum has higher energy in the high-frequency part. The energy spectrum of an image can be obtained by calculating the Fourier transform of the image, and the clarity of the image can be evaluated based on this.

[0052] In the present application, in the above step 303, based on the image of the feature to be recognized at the first brightness in the nth brightness interval, using a band-pass algorithm, determine the clarity of the image of the feature to be recognized at the first brightness in the nth brightness interval, specifically including:

[0053] Based on the image of the feature to be recognized at the first brightness in the nth brightness interval, use the fast Fourier transform to obtain the frequency response curve of the image of the feature to be recognized at the first brightness in the nth brightness interval.

[0054] Integrate the frequency response curve within a set frequency range to obtain the total energy of the image of the feature to be recognized at the first brightness in the nth brightness interval within the set frequency range.

[0055] Normalize the total energy within the set frequency range to obtain the clarity of the image of the feature to be recognized at the first brightness in the nth brightness interval.

[0056] Specifically, in practical applications, a band-pass filter is used to set the frequency range. A band-pass filter is a frequency-domain filter used to pass information within a specific frequency range by selectively retaining it while suppressing signals of other frequencies. It is usually used to remove specific frequency noise in an image or enhance specific frequency components.

[0057] The implementation method of the band-pass filter in the present application is:

[0058] .

[0059] Among them, is the filter response at the frequency domain midpoint point, is the distance from the midpoint in the frequency domain to the center, is the standard deviation of the Gaussian distribution, is the radial center of the bandwidth.

[0060] The calculation formula for the total energy within the set frequency range in this application is:

[0061] .

[0062] Among them, is the spectral representation of the signal, and are the two cut-off frequencies of the filter respectively, is the total energy within the set frequency range.

[0063] The calculation formula for normalization in this application is:

[0064] .

[0065] Among them, is the normalized sharpness.

[0066] In another exemplary embodiment of this application, step 203 specifically includes: obtaining the feature image to be recognized under the optimal light source and optimal brightness to obtain the optimal image.

[0067] Based on the optimal image, using the autocorrelation algorithm and the extreme value search algorithm, determine the template area of the optimal image. Specifically, it is the following steps 401 to step 404.

[0068] Step 401, based on the optimal image, use the autocorrelation algorithm to obtain the autocorrelation image of the optimal image.

[0069] Among them, the autocorrelation of the autocorrelation algorithm refers to the degree of correlation between a signal (or image) and its own copies at different times or spatial positions. In image processing, autocorrelation is usually used to measure the similarity or repeatability between a certain area in an image and its adjacent areas.

[0070] In order to determine the difference or similarity degree between the image signal and , it is usually represented by the autocorrelation function:

[0071] .

[0072] Among them, is the autocorrelation function, is the image signal at time is the image signal at time is the offset.

[0073] Characteristics of the autocorrelation function: The autocorrelation function is an even function , when is 0, the autocorrelation function is equal to the energy of the signal is the maximum value of the autocorrelation function.

[0074] In image processing, the steps to obtain the autocorrelation image are as follows: First, Fourier transform. By performing a Fourier transform on the input image, it is converted from the spatial domain to the frequency domain. Second, frequency-domain image calculation. Based on the two obtained frequency-domain images, the two frequency-domain images are multiplied point by point to obtain a convolution image. Third, inverse Fourier transform. Based on the convolution image, it is converted back to the spatial domain through an inverse Fourier transform to obtain the autocorrelation image of the input image.

[0075] Step 402: Based on the autocorrelation image of the optimal image, use an extreme value search algorithm to obtain the positions of multiple final feature points in the autocorrelation image.

[0076] Step 403: Based on the positions of multiple final feature points in the autocorrelation image, determine the central position of the autocorrelation image.

[0077] Step 404: Based on the central position of the autocorrelation image and the positions of four adjacent points, determine the template area of the optimal image.

[0078] In another exemplary embodiment of the present application, step 402 specifically includes: Based on the autocorrelation image of the optimal image, obtain multiple pixel maximum points of the autocorrelation image.

[0079] Based on the multiple pixel maximum points, use a sub-pixel interpolation method to obtain the positions of multiple feature points.

[0080] Based on the multiple pixel maximum points and a set threshold, screen the multiple feature points to obtain the positions of multiple final feature points in the autocorrelation image.

[0081] In another exemplary embodiment of the present application, the traditional way to change the camera's field of view is: a small target surface camera is combined with a zoom lens, and the field of view is modified by controlling the lens to change the magnification. This method has the problem of too low lens resolution at high magnifications.

[0082] In this application, the field of view of the camera can be changed by combining a large target surface camera (2448*2048 pixel) with multiple groups of fixed magnification lenses, by cutting the resolution of the camera itself. And the switching of the field of view does not affect the accuracy of the vision system itself and the lens resolution ability, improving the overall vision accuracy. At the same time, through pre-calibration, the camera is calibrated under the condition of full resolution. Even if the resolution is cut, there is no need to re-calibrate, improving the switching efficiency. And this application can achieve stepless transformation. By automatically obtaining the size of the wafer or the feature to be recognized, the field of view is automatically calculated to obtain a suitable resolution, without waste of the field of view.

[0083] In step 204 of this application, it specifically includes: determining the sizes of the sides of the template area based on the template area of the optimal image.

[0084] Determine the region of interest of the optimal image with a set multiple of the sides of the template area.

[0085] In another exemplary embodiment of this application, step 205 specifically includes: determining the length and width of the region of interest based on the size of the region of interest of the optimal image.

[0086] Based on the length and width of the region of interest, use the calibrated camera to crop the optimal image to obtain the region of interest image of the feature to be recognized.

[0087] This application improves the stability and quality of the mounting position and angle accuracy, enhances the efficiency during visual replacement of different varieties, reduces the error of manual operation, and ensures the quality and performance of semiconductor products through the vision intelligent adjustment method of a fully automatic mounter.

[0088] Based on the same inventive concept, as Figure 5 shown, the embodiment of this application also provides a vision intelligent adjustment system for a fully automatic mounter. It includes:

[0089] An image acquisition module 501, configured to collect an image of the feature to be recognized in real time based on the camera in the fully automatic mounter to obtain a real-time image; the feature to be recognized is a wafer or a substrate.

[0090] An optimal light source and optimal brightness determination module 502, connected to the image acquisition module 501, for determining the optimal light source and optimal brightness of the feature to be recognized based on multiple light sources in the fully automatic mounter and the real-time image.

[0091] A template area determination module 503, connected to the optimal light source and optimal brightness determination module 502, for obtaining the image of the feature to be recognized under the optimal light source and optimal brightness to obtain an optimal image, and determining the template area of the optimal image.

[0092] The region of interest determination module 504, which is connected to the template region determination module 503, is configured to determine the region of interest of the optimal image based on the template region of the optimal image.

[0093] The adjustment module 505, which is connected to the region of interest determination module 504, is configured to crop the optimal image based on the region of interest of the optimal image to obtain an image of interest for the feature to be recognized; and adjust the camera to be recognized based on the image of interest for the feature to be recognized.

[0094] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 6 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store video tag processing data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a method for visual intelligent adjustment of a fully automatic chip mounter.

[0095] Those skilled in the art can understand that Figure 6 the structure shown in

[0096] is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0097] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0098] The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0099] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0100] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A visual intelligent adjustment method for a fully automatic wafer loading machine, characterized in that: The visual intelligent adjustment method of the fully automatic loader comprises: Based on the camera in the fully automatic wafer mounter, an image of the feature to be identified is collected in real time to obtain a real-time image; the feature to be identified is a wafer or a substrate; Based on the multiple light sources in the fully automatic loader and the real-time image, the optimal light source and the optimal brightness of the feature to be identified are determined, specifically including: Based on the multiple light sources in the fully automatic loader and the real-time image, the optimal light source and the optimal brightness of the feature to be identified are determined by using the golden section search algorithm and the bandpass algorithm, specifically including: For the nth iteration of any light source, based on the nth brightness interval of the light source, a golden section search algorithm is used to determine a first brightness in the nth brightness interval and a second brightness in the nth brightness interval; n>0; Acquire a feature image to be identified at a first brightness in an nth brightness interval and a feature image to be identified at a second brightness in an nth brightness interval; Based on the feature image to be identified at the first brightness in the nth brightness interval and the feature image to be identified at the second brightness in the nth brightness interval, a bandpass algorithm is used to determine the clarity of the feature image to be identified at the first brightness in the nth brightness interval and the clarity of the feature image to be identified at the second brightness in the nth brightness interval; Based on the definition, determining an n+1th brightness interval according to an nth brightness interval, a first brightness in the nth brightness interval, and a second brightness in the nth brightness interval; Determine whether n is less than the total number of iterations, and if so, perform the n+1th iteration; if not, determine the maximum clarity value of the light source based on the clarity of the feature image to be identified at the first brightness and the clarity of the feature image to be identified at the second brightness in the first brightness interval to the nth brightness interval; Determining the optimal light source and optimal brightness of the feature to be identified based on the maximum clarity values ​​of the multiple light sources; Acquire the feature image to be identified under the optimal light source and optimal brightness, obtain the optimal image, and determine the template area of ​​the optimal image, specifically including: Acquire the feature image to be identified under the optimal light source and optimal brightness to obtain the optimal image; Based on the optimal image, the template area of ​​the optimal image is determined by using the autocorrelation algorithm and the extreme value search algorithm; Based on the template area of ​​the optimal image, the region of interest of the optimal image is determined, specifically including: Based on the template area of ​​the optimal image, determine the length of each side of the template area; Determine the region of interest of the optimal image by setting multiples of the lengths of each side of the template region; Based on the region of interest of the optimal image, the optimal image is cropped to obtain an image of interest of the feature to be identified; and the camera is adjusted based on the image of interest of the feature to be identified.

2. The visual intelligent adjustment method of the fully automatic wafer loading machine according to claim 1, characterized in that: Based on the feature image to be identified at the first brightness in the nth brightness interval, a bandpass algorithm is used to determine the clarity of the feature image to be identified at the first brightness in the nth brightness interval, specifically including: Based on the feature image to be identified at the first brightness in the nth brightness interval, a frequency response curve of the feature image to be identified at the first brightness in the nth brightness interval is obtained by using a fast Fourier transform; Integrate the frequency response curve within a set frequency range to obtain the total energy of the feature image to be identified at the first brightness in the nth brightness interval within the set frequency range; The total energy within the set frequency range is normalized to obtain the clarity of the feature image to be identified at the first brightness in the nth brightness interval.

3. The visual intelligent adjustment method of the fully automatic loader according to claim 1, characterized in that: Based on the optimal image, the template area of ​​the optimal image is determined using the autocorrelation algorithm and the extreme value search algorithm, including: Based on the optimal image, an autocorrelation algorithm is used to obtain an autocorrelation image of the optimal image; Based on the autocorrelation image of the optimal image, an extreme value search algorithm is used to obtain the positions of multiple final feature points in the autocorrelation image; Determine the center position of the autocorrelation image based on the positions of the multiple final feature points in the autocorrelation image; Based on the center position of the autocorrelation image and the positions of four adjacent points, the template area of ​​the optimal image is determined.

4. The visual intelligent adjustment method of the fully automatic wafer loading machine according to claim 3 is characterized in that: Based on the autocorrelation image of the optimal image, an extreme value search algorithm is used to obtain the positions of multiple final feature points in the autocorrelation image, including: Based on the autocorrelation image of the optimal image, multiple pixel maximum points of the autocorrelation image are obtained; Based on multiple pixel maximum points, the sub-pixel interpolation method is used to obtain the positions of multiple feature points; Based on multiple pixel maximum points and set thresholds, multiple feature points are screened to obtain the positions of multiple final feature points in the autocorrelation image.

5. The visual intelligent adjustment method of the fully automatic loader according to claim 1, characterized in that: Based on the region of interest of the optimal image, the optimal image is cropped to obtain an image of interest of the feature to be identified, specifically including: Determine the length and width of the region of interest based on the size of the region of interest of the optimal image; Based on the length and width of the region of interest, the optimal image is cropped using a calibrated camera to obtain an image of interest with features to be identified.

6. A visual intelligent adjustment system for a fully automatic wafer loading machine, using the visual intelligent adjustment method for a fully automatic wafer loading machine according to any one of claims 1 to 5, characterized in that: The visual intelligent adjustment system of the fully automatic loader includes: An image acquisition module is used to acquire an image of a feature to be identified in real time based on a camera in a fully automatic wafer mounter to obtain a real-time image; the feature to be identified is a wafer or a substrate; An optimal light source and optimal brightness determination module, connected to the image acquisition module, for determining the optimal light source and optimal brightness of the feature to be identified based on the multiple light sources in the full-automatic loader and the real-time image; A template region determination module, connected to the optimal light source and optimal brightness determination module, is used to obtain a feature image to be identified under the optimal light source and optimal brightness, obtain an optimal image, and determine a template region of the optimal image; An area of ​​interest determination module, connected to the template area determination module, and configured to determine the area of ​​interest of the optimal image based on the template area of ​​the optimal image; The adjustment module is connected to the region of interest determination module and is used to crop the optimal image based on the region of interest of the optimal image to obtain an image of interest of the feature to be identified; and adjust the camera based on the image of interest of the feature to be identified.

Images