CN117463645A - Automatic control method and system for semiconductor sorting integrated machine - Google Patents

Automatic control method and system for semiconductor sorting integrated machine Download PDF

Info

Publication number
CN117463645A
CN117463645A CN202311823941.0A CN202311823941A CN117463645A CN 117463645 A CN117463645 A CN 117463645A CN 202311823941 A CN202311823941 A CN 202311823941A CN 117463645 A CN117463645 A CN 117463645A
Authority
CN
China
Prior art keywords
semiconductor
image
gripper
coordinate
grabbing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202311823941.0A
Other languages
Chinese (zh)
Other versions
CN117463645B (en
Inventor
叶十逢
俞子强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Dongyi Semiconductor Technology Jiangsu Co ltd
Original Assignee
Dongyi Semiconductor Technology Jiangsu Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Dongyi Semiconductor Technology Jiangsu Co ltd filed Critical Dongyi Semiconductor Technology Jiangsu Co ltd
Priority to CN202311823941.0A priority Critical patent/CN117463645B/en
Publication of CN117463645A publication Critical patent/CN117463645A/en
Application granted granted Critical
Publication of CN117463645B publication Critical patent/CN117463645B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/36—Sorting apparatus characterised by the means used for distribution
    • B07C5/361—Processing or control devices therefor, e.g. escort memory
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/36—Sorting apparatus characterised by the means used for distribution
    • B07C5/361—Processing or control devices therefor, e.g. escort memory
    • B07C5/362—Separating or distributor mechanisms
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C2501/00—Sorting according to a characteristic or feature of the articles or material to be sorted
    • B07C2501/0063—Using robots
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Container, Conveyance, Adherence, Positioning, Of Wafer (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides an automatic control method and system for a semiconductor sorting integrated machine, which relate to the technical field of data processing, and are characterized in that a relevant gripper position mark is generated by receiving a semiconductor image to be sorted and a reference semiconductor image identification so as to call a plurality of vacuum chuck grippers to execute grabbing positioning control and collect gripper positioning images, further, a gripper coordinate deviation coefficient is determined based on sorting gripper positioning image information, and when the deviation coefficient is smaller than or equal to a coordinate deviation coefficient threshold value, the plurality of vacuum chuck grippers are activated to execute semiconductor sorting. The technical problem that the traditional semiconductor sorting machine cannot automatically generate grabbing schemes of multiple types of semiconductor chips due to the fact that the grabbing positions of the semiconductor chips are different in the prior art is solved. The technical effects of determining the grabbing site information for carrying out nondestructive grabbing of the semiconductor according to the automatic analysis of the shape characteristics of the semiconductor and improving the grabbing flexibility of the semiconductor sorting all-in-one machine for carrying out unspecified semiconductor chips are achieved.

Description

Automatic control method and system for semiconductor sorting integrated machine
Technical Field
The invention relates to the technical field of data processing, in particular to an automatic control method and system for a semiconductor sorting integrated machine.
Background
Due to the variety of semiconductor devices and the large differences in shape and size, different gripper designs and different gripping positions are required, which makes the scene very complex, and the traditional sorting machine is difficult to adapt to the variety, so that the automation requirement for gripping different types of semiconductor chips cannot be met.
In the prior art, the technical problem that the traditional semiconductor sorting machine cannot automatically generate the grabbing schemes of multiple types of semiconductor chips due to the difference of grabbing positions based on different types of semiconductor chips.
Disclosure of Invention
The application provides an automatic control method and an automatic control system for a semiconductor sorting integrated machine, which are used for solving the technical problem that the traditional semiconductor sorting machine cannot realize automatic generation of a grabbing scheme of multiple types of semiconductor chips due to the fact that the differences based on grabbing positions of different types of semiconductor chips exist in the prior art.
In view of the above, the present application provides an automatic control method and system for a semiconductor sorting integrated machine.
In a first aspect of the present application, an automatic control method for a semiconductor sorting integrated machine is provided, the method comprising: when the semiconductor to be sorted is conveyed to a first platform of the semiconductor sorting integrated machine, an image sensor of a visual detection module is activated to collect images of the semiconductor to be sorted; configuring a preset semiconductor image library, wherein any reference semiconductor image of the preset semiconductor image library is provided with a gripper position mark, and a reference coordinate system of the gripper position mark is arranged on the reference semiconductor image; based on the preset semiconductor image library, activating a grabbing control identification module, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, and generating an associated gripper position identification; according to the position identification of the related grippers, a plurality of vacuum chuck grippers are called to execute grabbing positioning control, an image sensor is called after positioning is completed, and gripper positioning image information is acquired; activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information, and generating a gripper positioning coordinate mark; activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate mark and the associated gripper position mark, and generating a gripper coordinate deviation coefficient; and when the coordinate deviation coefficient of the grippers is smaller than or equal to the coordinate deviation coefficient threshold value, activating the plurality of vacuum chuck grippers to perform semiconductor sorting.
In a second aspect of the present application, there is provided an automatic control system for a semiconductor sorting integrated machine, the system comprising: the image acquisition and activation unit is used for activating an image sensor of the visual detection module to acquire images of the semiconductors to be sorted when the semiconductors to be sorted are conveyed to a first platform of the semiconductor sorting integrated machine; an image configuration execution unit, configured to configure a preset semiconductor image library, wherein any reference semiconductor image of the preset semiconductor image library has a gripper position identifier, and a reference coordinate system of the gripper position identifier is disposed in the reference semiconductor image; the identification module activation unit is used for activating a grabbing control identification module based on the preset semiconductor image library, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, and generating an associated gripper position identification; the positioning image acquisition unit is used for calling a plurality of vacuum chuck grippers to execute grabbing positioning control according to the position identification of the associated grippers, calling an image sensor after positioning is completed, and acquiring gripper positioning image information; the image information analysis unit is used for activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information and generating a gripper positioning coordinate mark; the position identification analysis unit is used for activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate identification and the associated gripper position identification and generating a gripper coordinate deviation coefficient; and the gripper activation execution unit is used for activating the plurality of vacuum chuck grippers to execute semiconductor sorting when the gripper coordinate deviation coefficient is smaller than or equal to the coordinate deviation coefficient threshold value.
One or more technical solutions provided in the present application have at least the following technical effects or advantages:
according to the method, when the semiconductor to be sorted is conveyed to the first platform of the semiconductor sorting integrated machine, the image sensor of the visual detection module is activated to collect the image of the semiconductor to be sorted; configuring a preset semiconductor image library, wherein any reference semiconductor image of the preset semiconductor image library is provided with a gripper position mark, and a reference coordinate system of the gripper position mark is arranged on the reference semiconductor image; based on the preset semiconductor image library, activating a grabbing control identification module, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, and generating an associated gripper position identification; according to the position identification of the related grippers, a plurality of vacuum chuck grippers are called to execute grabbing positioning control, an image sensor is called after positioning is completed, and gripper positioning image information is acquired; activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information, and generating a gripper positioning coordinate mark; activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate mark and the associated gripper position mark, and generating a gripper coordinate deviation coefficient; and when the coordinate deviation coefficient of the grippers is smaller than or equal to the coordinate deviation coefficient threshold value, activating the plurality of vacuum chuck grippers to perform semiconductor sorting. The method achieves the technical effects of indirectly improving the sorting efficiency of the semiconductor while improving the flexibility of the semiconductor sorting integrated machine for gripping the unspecified semiconductor chips according to the automatic analysis of the shape characteristics of the semiconductor to determine the information of the gripping sites for carrying out nondestructive gripping of the semiconductor.
Drawings
Fig. 1 is a schematic flow chart of an automatic control method for a semiconductor sorting integrated machine provided by the application;
fig. 2 is a schematic flow chart of configuring a preset semiconductor image library in the automatic control method for a semiconductor sorting integrated machine provided by the present application;
fig. 3 is a schematic structural diagram of an automatic control system for a semiconductor sorting integrated machine provided by the present application.
Reference numerals illustrate: the device comprises an image acquisition and activation unit 1, an image configuration execution unit 2, an identification module activation unit 3, a positioning image acquisition unit 4, an image information analysis unit 5, a position identification analysis unit 6 and a gripper activation execution unit 7.
Detailed Description
The application provides an automatic control method and an automatic control system for a semiconductor sorting integrated machine, which are used for solving the technical problem that the traditional semiconductor sorting machine cannot realize automatic generation of a grabbing scheme of multiple types of semiconductor chips due to the fact that the differences based on grabbing positions of different types of semiconductor chips exist in the prior art. The method achieves the technical effects of indirectly improving the sorting efficiency of the semiconductor while improving the flexibility of the semiconductor sorting integrated machine for gripping the unspecified semiconductor chips according to the automatic analysis of the shape characteristics of the semiconductor to determine the information of the gripping sites for carrying out nondestructive gripping of the semiconductor.
The technical scheme of the invention accords with related regulations on data acquisition, storage, use, processing and the like.
In the following, the technical solutions of the present invention will be clearly and completely described with reference to the accompanying drawings, and it should be understood that the described embodiments are only some embodiments of the present invention, but not all embodiments of the present invention, and that the present invention is not limited by the exemplary embodiments described herein. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention. It should be further noted that, for convenience of description, only some, but not all of the drawings related to the present invention are shown.
Example 1
As shown in fig. 1, the present application provides an automatic control method for a semiconductor sorting machine, which is applied to an automatic control system of the semiconductor sorting machine, wherein the system is embedded in the semiconductor sorting machine, the semiconductor sorting machine is provided with a plurality of vacuum chuck grippers, and the system comprises a grabbing control identification module and a visual detection module, and comprises:
specifically, in the present embodiment, the semiconductor sorting machine is a device integrating the semiconductor sorting function, has an integrated design, and can complete all steps of semiconductor sorting in the device of the semiconductor sorting machine, thereby improving efficiency, reducing cost, and ensuring uniform quality control.
Before the semiconductor sorting machine is used for separating defective products from good products in semiconductor chips, the in-machine vacuum chuck grippers are required to be used for carrying out nondestructive grabbing of the semiconductor chips, and the positions of the different semiconductor chips which can be subjected to nondestructive grabbing are different based on the diversity of the semiconductor chip design.
The embedded automatic control system of semiconductor sorting all-in-one, the system includes snatch control identification module and visual detection module, and this embodiment elaborates in the following description snatch control identification module and visual detection module's function and specific application scenario.
The automatic control system runs an automatic control method, and based on the automatic control method, personalized grabbing and sorting of different semiconductor devices can be achieved based on the semiconductor sorting integrated machine, and the method is described in detail in the following description.
A100, when the semiconductor to be sorted is conveyed to a first platform of the semiconductor sorting integrated machine, an image sensor of a visual detection module is activated to collect images of the semiconductor to be sorted;
Specifically, in this embodiment, the semiconductor to be sorted is an unspecified semiconductor chip for mass production, and the first platform is located at a machine inlet of the semiconductor sorting integrated machine, and is configured to provide suitable illumination to facilitate the image sensor of the vision detection module to perform image acquisition with high reduction degree on the semiconductor to be analyzed.
When the semiconductor to be sorted is conveyed to a first platform of the semiconductor sorting integrated machine, an image sensor of a visual detection module is activated to collect images of the semiconductor to be sorted, the image collecting sensor is preferably arranged at the upper end of the central position of the first platform, and the images of the semiconductor to be sorted obtained based on the image sensor of the visual detection module are top views of the semiconductor to be sorted.
A200, configuring a preset semiconductor image library, wherein any reference semiconductor image of the preset semiconductor image library is provided with a hand grip position mark, and a reference coordinate system of the hand grip position mark is arranged on the reference semiconductor image;
in one embodiment, as shown in fig. 2, a preset semiconductor image library is configured, where any reference semiconductor image of the preset semiconductor image library has a gripper position identifier, where a reference coordinate system of the gripper position identifier is disposed on the reference semiconductor image, and a method step a200 provided in the present application includes:
A210, obtaining a first semiconductor of a preset semiconductor set, and constructing a first semiconductor digital image;
a220, constructing a first reference coordinate system in the first semiconductor digital image by taking the central position of the first semiconductor digital image as the origin of the coordinate system;
a230, combining the first reference coordinate system, carrying out grabbing configuration optimization on the first semiconductor, generating a first semiconductor gripper position identifier, storing the first semiconductor gripper position identifier and the first semiconductor digital image in a correlated mode, and adding the first semiconductor gripper position identifier and the first semiconductor digital image into the preset semiconductor image library.
In one embodiment, in combination with the first reference coordinate system, performing grasping configuration optimization on the first semiconductor to generate a first semiconductor gripper location identifier, a method step a230 provided in the present application further includes:
a231, obtaining first semiconductor basic information, wherein the first semiconductor basic information comprises first semiconductor weight information and a first semiconductor plane area position mark;
a232, loading the rated weight of a single gripper, and matching the constraint quantity of the grippers by combining the first semiconductor weight information;
a233, randomly distributing position identifiers in the first semiconductor plane area according to the constraint quantity of the grippers to generate a grabbing position assignment result;
A234, carrying out distribution uniformity evaluation on the grasping position assignment result to generate a distribution uniformity coefficient;
a235, setting the grabbing position assignment result as the first semiconductor gripper position mark when the distribution uniformity coefficient meets a distribution uniformity coefficient threshold value.
In one embodiment, the evaluation of distribution uniformity is performed on the assignment result of the grabbing position, so as to generate a distribution uniformity coefficient, and the method step a234 provided in the present application further includes:
a234-1, carrying out position-by-position distance analysis on the grabbing position assignment result to generate a distribution distance feature set;
a234-2, carrying out edge position connection on the grabbing position assignment result, and counting area characteristics of a delineating area;
a234-3, performing variance calculation based on the distribution distance feature set to generate a first distribution uniformity coefficient;
a234-4, calculating the ratio of the area characteristic of the delineating area to the total area of the first semiconductor, and setting the ratio as a second distribution uniformity coefficient;
and A234-5, adding the first distribution uniformity coefficient and the second distribution uniformity coefficient into the distribution uniformity coefficient.
Specifically, in this embodiment, the preset semiconductor image library is a database for storing hand position identifiers of semiconductor gripping positions of multiple types of standard semiconductors, and a reference coordinate system of the hand position identifiers is disposed in a reference semiconductor image.
The generation process of the preferable configuration of the preset semiconductor image library is as follows, the preset semiconductor set comprising the multi-type reference semiconductors is obtained interactively, the first semiconductor is obtained based on random calling of the preset semiconductor set, and the first semiconductor is an unspecified type semiconductor chip.
And interactively obtaining first semiconductor etching design information, carrying out image reduction based on the first semiconductor etching design information, obtaining the first semiconductor digital image of the appearance image of the reference semiconductor which is theoretically under the condition of qualified production, and constructing a first reference coordinate system in the first semiconductor digital image by taking the central position of the first semiconductor digital image as the origin of the coordinate system.
And combining the first reference coordinate system, performing grabbing configuration optimization on the first semiconductor to obtain the first semiconductor gripper position identification representing the surface positions of a plurality of semiconductor chips for performing nondestructive grabbing on the first semiconductor.
The process of grasping configuration optimizing specifically comprises the following steps:
and interacting the design information of the first semiconductor to obtain the first semiconductor basic information, wherein the first semiconductor basic information specifically comprises first semiconductor weight information and a first semiconductor plane area position identifier, and the plane area position identifier identifies all the flat areas which are not etched in the device of the first semiconductor.
And interacting machine design information of the semiconductor sorting integrated machine so as to load and obtain rated weight of the single gripper, wherein the rated weight represents the maximum weight or load which each vacuum chuck gripper can safely bear under normal working conditions.
Taking the first semiconductor weight information as a dividend, taking the single-gripper rated weight as a divisor, and carrying the calculated decimal (the number of gripper constraints=semiconductor weight/single-gripper rated weight, if the number of gripper constraints is 4.5, carrying is 5), so as to obtain the number of gripper constraints representing the number of vacuum chuck grippers required for stably gripping the first semiconductor, wherein the number of gripper constraints is specifically H vacuum chuck grippers.
And randomly distributing the position identifiers of the first semiconductor plane area to meet the H number of the grabbing sites of the grippers in the constraint quantity of the grippers, and taking the H number of the grabbing sites as the grabbing position assignment result.
Performing a two-by-two gripper gripping site position distance analysis of the H gripper gripping sites on the gripping position assignment result to generate a gripping position assignment result, wherein the gripping position assignment result comprisesAnd the distribution distance feature set of the position distances of the grabbing sites of each grabbing hand, wherein each grabbing hand grabbing site position distance corresponds to two grabbing site positions.
And selecting a gripper grabbing site positioned at the edge of the scattered point from the H gripper grabbing sites of the grabbing position assignment result to carry out edge position connection, so as to obtain a circled area containing the H gripper grabbing sites, and further counting the area characteristics of the circled area based on the existing image area determination method, wherein the area characteristics of the circled area are the area data of the circled area.
And carrying out variance calculation based on the distribution distance feature set to generate a first distribution uniformity coefficient, wherein the smaller the numerical value of the first distribution uniformity coefficient is, the closer the position distances of a plurality of groups of hand grabbing sites in the distribution distance feature set are consistent, and the smaller the numerical value of the first distribution uniformity coefficient is, the better the position distances of a plurality of groups of hand grabbing sites in the distribution distance feature set are. And calculating the ratio of the area characteristic of the delineating region in the total area of the first semiconductor, and setting the ratio as a second distribution uniformity coefficient, wherein the larger the value of the second distribution uniformity coefficient is, the higher the dispersion degree of the H grabbing sites on the surface of the first semiconductor is, and the larger the value of the second distribution uniformity coefficient is. And adding the first distribution uniformity coefficient and the second distribution uniformity coefficient into the distribution uniformity coefficient.
The method comprises the steps of presetting a distribution uniformity coefficient threshold, wherein the distribution uniformity coefficient threshold specifically comprises a first distribution uniformity coefficient threshold and a second distribution uniformity coefficient threshold, and when the first distribution uniformity coefficient is smaller than the first distribution uniformity coefficient threshold and the second distribution uniformity coefficient is larger than the second distribution uniformity coefficient threshold, the distribution uniformity coefficient is considered to meet the distribution uniformity coefficient threshold, and the grabbing position assignment result is set to be the first semiconductor gripper position identification.
And carrying out position location on the first semiconductor hand grip position identification in the first semiconductor digital image by combining a first reference coordinate system to finish the association storage, obtaining a first semiconductor digital image with H hand grip grabbing site identifications, and adding the first semiconductor digital image into the preset semiconductor image library.
And similarly, obtaining a plurality of groups of semiconductor gripper position identifiers and semiconductor digital images of a plurality of semiconductor chips in the preset semiconductor set by adopting the same method, and completing configuration of the preset semiconductor image library by adopting a mode of adding the associated storage into the preset semiconductor image library.
The purpose of constructing the preset semiconductor image library in this embodiment is to quickly position the gripping sites of the chip gripper by referring to the preset semiconductor image library after obtaining a semiconductor chip, so as to implement gripping control of the semiconductor chip by multiple vacuum chuck grippers of the semiconductor sorting integrated machine.
A300, activating a grabbing control identification module based on the preset semiconductor image library, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, and generating an associated gripper position identification;
In one embodiment, based on the preset semiconductor image library, a capture control identification module is activated, the semiconductor image to be sorted and the reference semiconductor image are received for identification, and an associated gripper position identification is generated, and the method step a300 provided in the present application further includes:
a310, activating a first feature analysis channel of the grabbing control identification module, and carrying out feature extraction on the semiconductor images to be sorted to generate a first convolution feature extraction result;
a320, activating a second characteristic analysis channel of the grabbing control identification module, and carrying out characteristic extraction on the reference semiconductor image to generate a second convolution characteristic extraction result, wherein the first characteristic analysis channel and the second characteristic analysis channel have the same model structure and model parameters;
a330, activating a feature comparison channel of the grabbing control identification module, and comparing the first convolution feature extraction result with the second convolution feature extraction result to generate a feature comparison result, wherein the feature comparison result comprises 0 or 1;
a340, when the characteristic comparison result is equal to 1, extracting a semiconductor gripper position identifier of the reference semiconductor image from the preset semiconductor image library, and setting the semiconductor gripper position identifier as the associated gripper position identifier;
And A350, updating the reference semiconductor image for comparison when the characteristic comparison result is equal to 0.
In one embodiment, based on the preset semiconductor image library, a capture control identification module is activated, and the semiconductor image to be sorted and the reference semiconductor image are received for identification, so as to generate an associated gripper position identification, before the method step a330 provided in the present application further includes:
a331, acquiring preset image acquisition parameters of an image sensor of a visual detection module, wherein the preset image acquisition parameters are used for acquiring a semiconductor image of a first platform;
a332, acquiring a semiconductor image data set and a semiconductor shape characteristic identification data set of a preset semiconductor type set by taking the preset image acquisition parameters as constraints, and training a characteristic analysis channel to be set as the first characteristic analysis channel and the second characteristic analysis channel;
a333, constructing a feature comparison function:
，
，
，
wherein A represents the semiconductor image global edge coordinate sequence, B represents the reference semiconductor image global edge coordinate sequence,any one pixel coordinate point for representing the whole edge coordinate sequence of the semiconductor image is +.>Any one pixel coordinate point of the whole edge coordinate sequence of the characteristic reference semiconductor image is +. >Representing the number of coincident pixels A and B in any edge alignment mode, +.>Characterizing the maximum number of overlapping pixels, < >>Characterizing the total number of global edge coordinate sequences of the reference semiconductor image,/->Characterization of A and B global edge shape similarity, +.>Characterizing semiconductor image local shape feature point coordinates, < >>Characterizing local shape feature point coordinates of the reference semiconductor image, < >>Characterizing the total number of feature points of the local shape of the reference semiconductor image, < >>Characterised in thatPartial shape similarity in alignment mode of +.>Characterizing the output value->Characterizing a first similarity threshold,/a>Characterizing a second similarity threshold;
a334, training the feature comparison channel according to the feature comparison function and combining a convolutional neural network.
Specifically, in this embodiment, preset image acquisition parameters of the image sensor of the vision detection module are obtained interactively, the preset image acquisition parameters are image sensor parameter set values for acquiring semiconductor images of the first platform, in step S100, when the semiconductor to be sorted is conveyed to the first platform of the semiconductor sorting all-in-one machine, the image sensor of the vision detection module is activated to acquire the semiconductor images to be sorted, and the semiconductor images to be sorted are acquired by image acquisition when the image sensor is adjusted to be the preset image acquisition parameters.
It should be understood that, due to the difference of the etching complexity of the semiconductor chip, the image acquisition accuracy of the semiconductor chip is limited by the etching complexity, so that the numerical value of the preset image acquisition parameter is not specifically limited in this embodiment, and the numerical value can be set according to the specific etching condition.
And debugging the image sensor by taking the preset image acquisition parameters as constraints, and further acquiring the images of the qualified real objects of the plurality of reference semiconductors in the preset semiconductor type set based on the image sensor to obtain the semiconductor image data set.
Further, the semiconductor shape feature identification dataset is obtained based on manually performing semiconductor feature region identification of each piece of semiconductor image data in the semiconductor image dataset.
And constructing a characteristic analysis channel based on the convolutional neural network, wherein the input data of the characteristic analysis channel is a semiconductor image, and the output result is a semiconductor shape characteristic identification image. Grouping the semiconductor image dataset and the semiconductor shape feature identification dataset into a plurality of sets of semiconductor image data-semiconductor shape feature identification data based on a reference semiconductor.
Dividing a plurality of groups of semiconductor image data-semiconductor shape characteristic identification data into a training set, a testing set and a verification set according to the data volume of 18:1:1, training and testing the semiconductor characteristic identification of the characteristic analysis channel based on the training set and the testing set, and verifying the accuracy of the semiconductor characteristic identification of the characteristic analysis channel based on the verification set until the accuracy and the integrity of the semiconductor characteristic identification of the characteristic analysis channel are stable to be higher than 99%.
And directly copying the characteristic analysis channels to obtain the first characteristic analysis channel and the second characteristic analysis channel which have the same model structure and model parameters and are arranged in parallel.
Further, the present embodiment constructs a feature comparison function, where the feature comparison function is as follows:
，
，
，
in the feature comparison function, A represents the whole edge coordinate sequence of the semiconductor image, B represents the whole edge coordinate sequence of the reference semiconductor image,any one pixel coordinate point for representing the whole edge coordinate sequence of the semiconductor image is +.>Any one pixel coordinate point of the whole edge coordinate sequence of the characteristic reference semiconductor image is +.>Representing the number of coincident pixels A and B in any edge alignment mode, +. >Characterizing the maximum number of overlapping pixels, < >>Characterizing the total number of global edge coordinate sequences of the reference semiconductor image,/->Characterization of A and B global edge shape similarity, +.>Characterizing semiconductor image local shape feature point coordinates, < >>Characterizing local shape feature point coordinates of the reference semiconductor image, < >>Characterizing the total number of feature points of the local shape of the reference semiconductor image, < >>Characterized in->Partial shape similarity in alignment mode of +.>Characterizing the output value->Characterizing a first similarity threshold,/a>A second similarity threshold is characterized.
And interactively obtaining a plurality of groups of solid semiconductors of various reference semiconductors in the preset semiconductor image library, further carrying out image acquisition based on an image sensor and carrying out feature identification based on a feature analysis channel, and obtaining a plurality of groups of solid semiconductor image data-solid semiconductor shape feature identification data.
And training the feature comparison channel by adopting the same method of training the feature analysis channel and combining a convolutional neural network according to the feature comparison function.
And arranging the first characteristic analysis channel and the second characteristic analysis channel in parallel, connecting the output ends of the two characteristic analysis channels with the input end of the characteristic comparison channel, and further adaptively configuring the input end and the output end of the characteristic comparison channel to complete the construction of the grabbing control identification module.
And activating a first characteristic analysis channel of the grabbing control identification module, and carrying out characteristic extraction on the semiconductor images to be sorted to generate a first convolution characteristic extraction result.
And activating a second characteristic analysis channel of the grabbing control identification module, randomly calling the reference semiconductor image of any reference semiconductor from the preset semiconductor image library, inputting the reference semiconductor image into the second characteristic analysis channel for characteristic extraction, and generating a second convolution characteristic extraction result.
And activating a feature comparison channel of the grabbing control identification module, and comparing the first convolution feature extraction result with the second convolution feature extraction result to generate a feature comparison result, wherein the feature comparison result comprises 0 or 1.
When the characteristic comparison result is equal to 1, extracting a semiconductor gripper position identifier of the reference semiconductor image from the preset semiconductor image library, and setting the semiconductor gripper position identifier as the associated gripper position identifier; otherwise, when the characteristic comparison result is equal to 0, randomly calling the reference semiconductor image of any reference semiconductor from the preset semiconductor image library, updating the reference semiconductor image for comparison, and repeating the cycle until the characteristic comparison result is equal to 1, extracting the semiconductor gripper position identification of the reference semiconductor image from the preset semiconductor image library based on the characteristic comparison result, and setting the semiconductor gripper position identification as the associated gripper position identification.
The technical effect of quickly identifying and determining the position of the semiconductor gripper for carrying out nondestructive grabbing on the semiconductor to be sorted after the semiconductor to be analyzed is obtained is achieved by constructing the multi-channel grabbing control identification module.
A400, according to the position identification of the associated gripper, a plurality of vacuum chuck grippers are called to execute grabbing positioning control, an image sensor is called after positioning is completed, and gripper positioning image information is acquired;
a500, activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information, and generating a gripper positioning coordinate mark;
specifically, in this embodiment, a first reference coordinate system is constructed in the first semiconductor digital image by the same method, a virtual reference coordinate system is constructed on the semiconductor surface to be analyzed, and then the location of the associated gripper position identifier on the grabbing position of the semiconductor surface to be sorted is performed based on the virtual reference coordinate system, so that a plurality of vacuum chuck grippers are called to execute grabbing location control, an image sensor is called after the location is completed, gripper location image information is acquired, and the gripper location image information is the distribution position image of the vacuum chuck grippers on the current semiconductor surface to be sorted.
The visual detection module is used for evaluating whether the grabbing position of the vacuum chuck grippers on the surface of the semiconductor to be sorted accords with the position identification of the associated grippers or not, and specifically comprises a gripper positioning coordinate channel and a coordinate comparison channel, and the output end of the gripper positioning coordinate channel is connected with the input end of the coordinate comparison channel.
And activating a gripper positioning coordinate channel of the visual detection module, constructing a coordinate system in an image similar to the first reference coordinate system by taking the center of a semiconductor to be sorted as a coordinate origin in gripper positioning image information, and carrying out vectorization description of the position of the vacuum chuck gripper based on the coordinate system to generate the gripper positioning coordinate mark.
S600, activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate mark and the associated gripper position mark, and generating a gripper coordinate deviation coefficient;
in one embodiment, the coordinate comparison channel of the visual detection module is activated, the gripper positioning coordinate identifier and the associated gripper position identifier are analyzed, and a gripper coordinate deviation coefficient is generated, and the method step a600 provided in the present application further includes:
And A610, comparing the gripper positioning coordinate mark with the associated gripper position mark to obtain a non-coincident position duty ratio, and setting the duty ratio as the gripper coordinate deviation coefficient.
Specifically, in this embodiment, the gripper positioning coordinate identifier and the associated gripper position identifier are compared through traversal, the non-overlapping position duty ratio of the non-overlapping position occupying the total number of positions of the gripper positioning coordinate identifier and the associated gripper position identifier is obtained, the non-overlapping position duty ratio is set as the gripper coordinate deviation coefficient, and the gripper coordinate deviation coefficient represents the deviation degree of the gripping position of the actual vacuum chuck gripper from the theoretical associated gripper position identifier.
A700, when the coordinate deviation coefficient of the gripper is smaller than or equal to the coordinate deviation coefficient threshold value, activating the plurality of vacuum chuck grippers to execute semiconductor sorting.
It should be appreciated that deviations from the theoretical position of the actual vacuum chuck gripper gripping the semiconductor surface are unavoidable mechanical errors, and that the gripping of the semiconductor can still be ensured to be a non-destructive gripping as long as the mechanical errors are controlled within a certain range.
Based on the above, the embodiment sets the coordinate deviation coefficient threshold value which is not set by a specific value based on the semiconductor design condition, and activates the plurality of vacuum chuck grippers to perform semiconductor sorting when the coordinate deviation coefficient of the grippers is smaller than or equal to the coordinate deviation coefficient threshold value, otherwise, performs position adjustment of the plurality of vacuum chuck grippers according to the associated gripper position identifier until the coordinate deviation coefficient of the grippers is smaller than or equal to the coordinate deviation coefficient threshold value.
According to the embodiment, the technical effects of indirectly improving the semiconductor sorting efficiency while improving the flexibility of the semiconductor sorting integrated machine for grabbing the unspecified semiconductor chips are achieved by automatically analyzing and determining the grabbing site information for carrying out nondestructive grabbing of the semiconductor according to the shape characteristics of the semiconductor.
Example two
Based on the same inventive concept as the automatic control method for the semiconductor sorting integrated machine in the foregoing embodiment, as shown in fig. 3, the present application provides an automatic control system for a semiconductor sorting integrated machine, wherein the system includes:
the image acquisition and activation unit 1 is used for activating an image sensor of the visual detection module to acquire images of the semiconductors to be sorted when the semiconductors to be sorted are conveyed to a first platform of the semiconductor sorting integrated machine;
an image configuration execution unit 2, configured to configure a preset semiconductor image library, where any reference semiconductor image of the preset semiconductor image library has a gripper position identifier, and a reference coordinate system of the gripper position identifier is disposed in the reference semiconductor image;
an identification module activating unit 3, configured to activate a capture control identification module based on the preset semiconductor image library, receive a semiconductor image to be sorted and a reference semiconductor image for identification, and generate an associated gripper position identification;
The positioning image acquisition unit 4 is used for calling a plurality of vacuum chuck grippers to execute grabbing positioning control according to the position identification of the associated grippers, calling an image sensor after positioning is completed, and acquiring gripper positioning image information;
the image information analysis unit 5 is used for activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information and generating a gripper positioning coordinate mark;
the position identification analysis unit 6 is used for activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate identification and the associated gripper position identification and generating a gripper coordinate deviation coefficient;
and the gripper activation execution unit 7 is used for activating the plurality of vacuum chuck grippers to execute semiconductor sorting when the coordinate deviation coefficient of the grippers is smaller than or equal to the coordinate deviation coefficient threshold value.
In one embodiment, the image configuration execution unit 2 further includes:
obtaining a first semiconductor of a preset semiconductor set, and constructing a first semiconductor digital image;
constructing a first reference coordinate system in the first semiconductor digital image by taking the central position of the first semiconductor digital image as the origin of the coordinate system;
And combining the first reference coordinate system, performing grabbing configuration optimization on the first semiconductor, generating a first semiconductor gripper position identifier, storing the first semiconductor gripper position identifier and the first semiconductor digital image in a correlated manner, and adding the first semiconductor gripper position identifier and the first semiconductor digital image into the preset semiconductor image library.
In one embodiment, the image configuration execution unit 2 further includes:
obtaining first semiconductor basic information, wherein the first semiconductor basic information comprises first semiconductor weight information and a first semiconductor plane area position mark;
loading the rated weight of a single gripper, and matching the constraint quantity of the grippers by combining the first semiconductor weight information;
randomly distributing position identifiers in the first semiconductor plane area according to the constraint quantity of the grippers to generate a grabbing position assignment result;
evaluating the distribution uniformity of the grasping position assignment result to generate a distribution uniformity coefficient;
and when the distribution uniformity coefficient meets a distribution uniformity coefficient threshold, setting the grabbing position assignment result as the first semiconductor gripper position mark.
In one embodiment, the image configuration execution unit 2 further includes:
performing position distance analysis on the grabbing position assignment result to generate a distribution distance feature set;
Performing edge position connection on the grabbing position assignment result, and counting area characteristics of the delineating area;
performing variance calculation based on the distribution distance feature set to generate a first distribution uniformity coefficient;
calculating the ratio of the area characteristic of the defined area in the total area of the first semiconductor, and setting the ratio as a second distribution uniformity coefficient;
and adding the first distribution uniformity coefficient and the second distribution uniformity coefficient into the distribution uniformity coefficient.
In one embodiment, the identification module activation unit 3 further comprises:
activating a first characteristic analysis channel of the grabbing control identification module, and carrying out characteristic extraction on the semiconductor images to be sorted to generate a first convolution characteristic extraction result;
activating a second characteristic analysis channel of the grabbing control identification module, carrying out characteristic extraction on the reference semiconductor image, and generating a second convolution characteristic extraction result, wherein the first characteristic analysis channel and the second characteristic analysis channel have the same model structure and model parameters;
activating a feature comparison channel of the grabbing control identification module, and comparing the first convolution feature extraction result with the second convolution feature extraction result to generate a feature comparison result, wherein the feature comparison result comprises 0 or 1;
When the characteristic comparison result is equal to 1, extracting a semiconductor gripper position identifier of the reference semiconductor image from the preset semiconductor image library, and setting the semiconductor gripper position identifier as the associated gripper position identifier;
and when the characteristic comparison result is equal to 0, updating the reference semiconductor image for comparison.
In one embodiment, the identification module activation unit 3 further comprises:
acquiring preset image acquisition parameters of an image sensor of a visual detection module, wherein the preset image acquisition parameters are used for acquiring a semiconductor image of a first platform;
taking the preset image acquisition parameters as constraints, acquiring a semiconductor image data set and a semiconductor shape characteristic identification data set of a preset semiconductor type set, and training a characteristic analysis channel, wherein the characteristic analysis channel is set as the first characteristic analysis channel and the second characteristic analysis channel;
and (3) constructing a feature comparison function:
，
，
，
wherein A represents the semiconductor image global edge coordinate sequence, B represents the reference semiconductor image global edge coordinate sequence,characterization of semiconductorsAny one pixel coordinate point of the whole image edge coordinate sequence,/->Any one pixel coordinate point of the whole edge coordinate sequence of the characteristic reference semiconductor image is +. >Representing the number of coincident pixels A and B in any edge alignment mode, +.>Characterizing the maximum number of overlapping pixels, < >>Characterizing the total number of global edge coordinate sequences of the reference semiconductor image,/->Characterization of A and B global edge shape similarity, +.>Characterizing semiconductor image local shape feature point coordinates, < >>Characterizing local shape feature point coordinates of the reference semiconductor image, < >>Characterizing the total number of feature points of the local shape of the reference semiconductor image, < >>Characterised in thatPartial shape similarity in alignment mode of +.>Characterizing the output value->Characterization of the first embodimentA similarity threshold, +_>Characterizing a second similarity threshold;
and training the feature comparison channel by combining a convolutional neural network according to the feature comparison function.
In one embodiment, the location identity analysis unit 6 further comprises:
and comparing the gripper positioning coordinate mark with the associated gripper position mark to obtain a non-coincident position duty ratio, and setting the duty ratio as the gripper coordinate deviation coefficient.
Any of the methods or steps described above may be stored as computer instructions or programs in various non-limiting types of computer memories, and identified by various non-limiting types of computer processors, thereby implementing any of the methods or steps described above.
Based on the above-mentioned embodiments of the present invention, any improvements and modifications to the present invention without departing from the principles of the present invention should fall within the scope of the present invention.

Claims (8)

1. The automatic control method for the semiconductor sorting all-in-one machine is characterized by being applied to an automatic control system of the semiconductor sorting all-in-one machine, wherein the system is embedded in the semiconductor sorting all-in-one machine and is provided with a plurality of vacuum chuck grippers, and the system comprises a grabbing control identification module and a visual detection module and comprises the following steps:
when the semiconductor to be sorted is conveyed to a first platform of the semiconductor sorting integrated machine, an image sensor of a visual detection module is activated to collect images of the semiconductor to be sorted;
configuring a preset semiconductor image library, wherein any reference semiconductor image of the preset semiconductor image library is provided with a gripper position mark, and a reference coordinate system of the gripper position mark is arranged on the reference semiconductor image;
based on the preset semiconductor image library, activating a grabbing control identification module, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, and generating an associated gripper position identification;
According to the position identification of the related grippers, a plurality of vacuum chuck grippers are called to execute grabbing positioning control, an image sensor is called after positioning is completed, and gripper positioning image information is acquired;
activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information, and generating a gripper positioning coordinate mark;
activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate mark and the associated gripper position mark, and generating a gripper coordinate deviation coefficient;
and when the coordinate deviation coefficient of the grippers is smaller than or equal to the coordinate deviation coefficient threshold value, activating the plurality of vacuum chuck grippers to perform semiconductor sorting.
2. The method of claim 1, wherein configuring a library of pre-set semiconductor images, wherein any reference semiconductor image of the library of pre-set semiconductor images has a gripper location identifier, wherein a reference coordinate system of the gripper location identifier is disposed on the reference semiconductor image, comprises:
obtaining a first semiconductor of a preset semiconductor set, and constructing a first semiconductor digital image;
constructing a first reference coordinate system in the first semiconductor digital image by taking the central position of the first semiconductor digital image as the origin of the coordinate system;
And combining the first reference coordinate system, performing grabbing configuration optimization on the first semiconductor, generating a first semiconductor gripper position identifier, storing the first semiconductor gripper position identifier and the first semiconductor digital image in a correlated manner, and adding the first semiconductor gripper position identifier and the first semiconductor digital image into the preset semiconductor image library.
3. The method of claim 2, wherein performing grip configuration optimization on the first semiconductor in conjunction with the first reference frame to generate a first semiconductor grip location identifier comprises:
obtaining first semiconductor basic information, wherein the first semiconductor basic information comprises first semiconductor weight information and a first semiconductor plane area position mark;
loading the rated weight of a single gripper, and matching the constraint quantity of the grippers by combining the first semiconductor weight information;
randomly distributing position identifiers in the first semiconductor plane area according to the constraint quantity of the grippers to generate a grabbing position assignment result;
evaluating the distribution uniformity of the grasping position assignment result to generate a distribution uniformity coefficient;
and when the distribution uniformity coefficient meets a distribution uniformity coefficient threshold, setting the grabbing position assignment result as the first semiconductor gripper position mark.
4. The method of claim 3, wherein evaluating the distribution uniformity of the grasping location assignment result to generate a distribution uniformity coefficient comprises:
performing position distance analysis on the grabbing position assignment result to generate a distribution distance feature set;
performing edge position connection on the grabbing position assignment result, and counting area characteristics of the delineating area;
performing variance calculation based on the distribution distance feature set to generate a first distribution uniformity coefficient;
calculating the ratio of the area characteristic of the defined area in the total area of the first semiconductor, and setting the ratio as a second distribution uniformity coefficient;
and adding the first distribution uniformity coefficient and the second distribution uniformity coefficient into the distribution uniformity coefficient.
5. The method of claim 1, wherein activating a capture control identification module based on the library of preset semiconductor images, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, generating an associated gripper location identification, comprises:
activating a first characteristic analysis channel of the grabbing control identification module, and carrying out characteristic extraction on the semiconductor images to be sorted to generate a first convolution characteristic extraction result;
Activating a second characteristic analysis channel of the grabbing control identification module, carrying out characteristic extraction on the reference semiconductor image, and generating a second convolution characteristic extraction result, wherein the first characteristic analysis channel and the second characteristic analysis channel have the same model structure and model parameters;
activating a feature comparison channel of the grabbing control identification module, and comparing the first convolution feature extraction result with the second convolution feature extraction result to generate a feature comparison result, wherein the feature comparison result comprises 0 or 1;
when the characteristic comparison result is equal to 1, extracting a semiconductor gripper position identifier of the reference semiconductor image from the preset semiconductor image library, and setting the semiconductor gripper position identifier as the associated gripper position identifier;
and when the characteristic comparison result is equal to 0, updating the reference semiconductor image for comparison.
6. The method of claim 5, wherein activating a capture control identification module based on the library of preset semiconductor images, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, generating an associated gripper location identification, comprises, before:
acquiring preset image acquisition parameters of an image sensor of a visual detection module, wherein the preset image acquisition parameters are used for acquiring a semiconductor image of a first platform;
Taking the preset image acquisition parameters as constraints, acquiring a semiconductor image data set and a semiconductor shape characteristic identification data set of a preset semiconductor type set, and training a characteristic analysis channel, wherein the characteristic analysis channel is set as the first characteristic analysis channel and the second characteristic analysis channel;
and (3) constructing a feature comparison function:
，
，
，
wherein A represents the semiconductor image global edge coordinate sequence, B represents the reference semiconductor image global edge coordinate sequence,any one pixel coordinate point for representing the whole edge coordinate sequence of the semiconductor image is +.>Any one pixel coordinate point of the whole edge coordinate sequence of the characteristic reference semiconductor image is +.>Representing the number of coincident pixels A and B in any edge alignment mode, +.>Characterizing the maximum number of overlapping pixels, < >>Characterizing the total number of global edge coordinate sequences of the reference semiconductor image,/->Characterization of A and B global edge shape similarity, +.>Characterizing semiconductor image local shape feature point coordinates, < >>Characterizing local shape feature point coordinates of the reference semiconductor image, < >>Characterizing the total number of feature points of the local shape of the reference semiconductor image, < >>Characterised in thatPartial shape similarity in alignment mode of +.>Characterizing the output value- >Characterizing a first similarity threshold,/a>Characterizing a second similarity threshold;
and training the feature comparison channel by combining a convolutional neural network according to the feature comparison function.
7. The method of claim 1, wherein activating the coordinate alignment channel of the visual detection module, analyzing the gripper location coordinate identification and the associated gripper location identification, generating gripper coordinate deviation coefficients, comprises:
and comparing the gripper positioning coordinate mark with the associated gripper position mark to obtain a non-coincident position duty ratio, and setting the duty ratio as the gripper coordinate deviation coefficient.
8. An automatic control system for a semiconductor sorting machine, the system comprising:
the image acquisition and activation unit is used for activating an image sensor of the visual detection module to acquire images of the semiconductors to be sorted when the semiconductors to be sorted are conveyed to a first platform of the semiconductor sorting integrated machine;
an image configuration execution unit, configured to configure a preset semiconductor image library, wherein any reference semiconductor image of the preset semiconductor image library has a gripper position identifier, and a reference coordinate system of the gripper position identifier is disposed in the reference semiconductor image;
The identification module activation unit is used for activating a grabbing control identification module based on the preset semiconductor image library, receiving a semiconductor image to be sorted and a reference semiconductor image for identification, and generating an associated gripper position identification;
the positioning image acquisition unit is used for calling a plurality of vacuum chuck grippers to execute grabbing positioning control according to the position identification of the associated grippers, calling an image sensor after positioning is completed, and acquiring gripper positioning image information;
the image information analysis unit is used for activating a gripper positioning coordinate channel of the visual detection module, analyzing the gripper positioning image information and generating a gripper positioning coordinate mark;
the position identification analysis unit is used for activating a coordinate comparison channel of the visual detection module, analyzing the gripper positioning coordinate identification and the associated gripper position identification and generating a gripper coordinate deviation coefficient;
and the gripper activation execution unit is used for activating the plurality of vacuum chuck grippers to execute semiconductor sorting when the gripper coordinate deviation coefficient is smaller than or equal to the coordinate deviation coefficient threshold value.
CN202311823941.0A 2023-12-28 2023-12-28 Automatic control method and system for semiconductor sorting integrated machine Active CN117463645B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311823941.0A CN117463645B (en) 2023-12-28 2023-12-28 Automatic control method and system for semiconductor sorting integrated machine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311823941.0A CN117463645B (en) 2023-12-28 2023-12-28 Automatic control method and system for semiconductor sorting integrated machine

Publications (2)

Publication Number Publication Date
CN117463645A true CN117463645A (en) 2024-01-30
CN117463645B CN117463645B (en) 2024-04-02

Family

ID=89624123

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311823941.0A Active CN117463645B (en) 2023-12-28 2023-12-28 Automatic control method and system for semiconductor sorting integrated machine

Country Status (1)

Country Link
CN (1) CN117463645B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2079633A (en) * 1980-05-14 1982-01-27 Lockwood Graders Ltd Article sorting apparatus and method
JP2012115785A (en) * 2010-12-02 2012-06-21 Sharp Corp Sorting system of waste
CN112338898A (en) * 2020-10-09 2021-02-09 中国矿业大学(北京) Image processing method and device of object sorting system and object sorting system
CN114332164A (en) * 2021-12-29 2022-04-12 北京半导体专用设备研究所（中国电子科技集团公司第四十五研究所） Disc-conveying semiconductor workpiece grabbing method and control system
US20230150777A1 (en) * 2020-04-03 2023-05-18 Beumer Group A/S Pick and place robot system, method, use and sorter system
CN116968022A (en) * 2023-07-14 2023-10-31 武汉纺织大学 Method and system for grabbing target object by mechanical arm based on visual guidance
CN117124332A (en) * 2023-10-17 2023-11-28 南京交通职业技术学院 Mechanical arm control method and system based on AI vision grabbing

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2079633A (en) * 1980-05-14 1982-01-27 Lockwood Graders Ltd Article sorting apparatus and method
JP2012115785A (en) * 2010-12-02 2012-06-21 Sharp Corp Sorting system of waste
US20230150777A1 (en) * 2020-04-03 2023-05-18 Beumer Group A/S Pick and place robot system, method, use and sorter system
CN112338898A (en) * 2020-10-09 2021-02-09 中国矿业大学(北京) Image processing method and device of object sorting system and object sorting system
CN114332164A (en) * 2021-12-29 2022-04-12 北京半导体专用设备研究所（中国电子科技集团公司第四十五研究所） Disc-conveying semiconductor workpiece grabbing method and control system
CN116968022A (en) * 2023-07-14 2023-10-31 武汉纺织大学 Method and system for grabbing target object by mechanical arm based on visual guidance
CN117124332A (en) * 2023-10-17 2023-11-28 南京交通职业技术学院 Mechanical arm control method and system based on AI vision grabbing

Also Published As

Publication number Publication date
CN117463645B (en) 2024-04-02

Similar Documents

Publication Publication Date Title
US8606536B2 (en) Methods and apparatus for hybrid outlier detection
US8788237B2 (en) Methods and apparatus for hybrid outlier detection
CN107085846B (en) Workpiece surface defect image identification method
CN104169945B (en) To the two-stage classification of the object in image
CN103776841B (en) Synthetic leather automatic defect detecting device and detection method
US8542917B2 (en) System and method for identifying complex tokens in an image
CN107944213A (en) The online Source Apportionments of PMF, system, terminal device and computer-readable recording medium
CN114066848A (en) FPCA appearance defect visual inspection system
TW202038110A (en) Classifying defects in a semiconductor specimen
CN117463645B (en) Automatic control method and system for semiconductor sorting integrated machine
CN115393645A (en) Automatic soil classification and naming method and system, storage medium and intelligent terminal
CN113269234B (en) Connecting piece assembly detection method and system based on target detection
CN112163799B (en) Yield analysis method and yield analysis system of semiconductor product
CN113933672A (en) Method and system for judging correlation of wafer test parameters
CN117245672A (en) Intelligent motion control system and method for modularized assembly of camera support
CN113033650A (en) Image classification method, training method and device of classification model and storage medium
CN115809622A (en) Chip simulation acceleration system with automatic optimization configuration function
CN103592304A (en) Stem object analysis system and stem object analysis method
CN113033469A (en) Tool damage identification method, device, equipment, system and readable storage medium
CN111222480A (en) Grape weight online estimation method and detection device based on deep learning
CN113642284B (en) Method and device for examining circuit schematic diagram, and computer readable storage medium
CN112328951A (en) Processing method of experimental data of analysis sample
CN116109627B (en) Defect detection method, device and medium based on migration learning and small sample learning
CN112348040B (en) Model training method, device and equipment
CN114817274B (en) Wafer data processing method and device, electronic device and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant