CN114965272A - Testing method of chip defect detection platform - Google Patents
Testing method of chip defect detection platform Download PDFInfo
- Publication number
- CN114965272A CN114965272A CN202210521687.8A CN202210521687A CN114965272A CN 114965272 A CN114965272 A CN 114965272A CN 202210521687 A CN202210521687 A CN 202210521687A CN 114965272 A CN114965272 A CN 114965272A
- Authority
- CN
- China
- Prior art keywords
- axis
- chip
- liquid lens
- slide rail
- axis direction
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8806—Specially adapted optical and illumination features
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
- G01N21/9501—Semiconductor wafers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
- G01N21/956—Inspecting patterns on the surface of objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/04—Batch operation; multisample devices
- G01N2201/0484—Computer controlled
Abstract
The application relates to a testing method of a chip defect detection platform, and relates to the field of chips. The method comprises the following steps: placing a chip to be tested on a chip clamping block; sending a first starting signal to a motion control card, wherein the motion control card controls an X-axis direction slide rail, a Y-axis direction slide rail and a Z-axis direction slide rail to move to corresponding preset coordinates; sending a second starting signal to the liquid lens control chip to control the liquid lens to automatically focus the chip to be tested; controlling a camera to take a picture to obtain an initial image in response to the fact that the liquid lens finishes automatic focusing on the chip to be tested; changing the numerical value of the preset Z-axis coordinate, and enabling the slide rail in the Z-axis direction to drive the liquid lens to move to the corresponding changed coordinate to obtain a plurality of groups of initial images; carrying out image post-processing on the multiple groups of initial images to obtain processed images; and inputting the processed image into a deep learning model for analysis to obtain a defect detection result of the chip to be detected. The method improves the appearance detection efficiency of the chip as a whole.
Description
Technical Field
The application relates to the technical field of chips, in particular to a testing method of a chip defect detection platform.
Background
Chips, often part of a computer or other electronic device. As a big chip consuming country in China, with the expansion of market demands and the upgrading of industrial scale, China has appeared a batch of brands with strong international competitiveness. Machine vision is very widely applied in the semiconductor industry, the application range is wider and wider, the detection and measurement of appearance defects, size, quantity, flatness, intervals, positioning, calibration, welding spot quality, bending degree and the like of a semiconductor are involved, and how to continuously, efficiently and quickly detect the appearance of a chip is a problem which needs to be solved urgently.
At present, common chip defect measuring methods mainly include manual detection, electron microscopy, ultrasonic scanning microscopy, machine vision and the like. (1) The traditional manual detection is time-consuming in statistical analysis, small in defect, difficult to locate and low in detectable rate, generally depends on manual subjective judgment, is low in consistency, limited in working time, easy to fatigue and difficult to deal with volatility. (2) The electron microscope mainly depends on density difference between different materials, and different material structure shadows are printed on the negative film after the chip is penetrated by rays, which is equivalent to a perspective top view, but double images can be generated for a logic chip with a more complex multi-layer structure, so that fine defects, gaps, cavity defects and the like on the back of the logic chip can not be detected. (3) The detection principle of the ultrasonic scanning microscope mainly utilizes the unique characteristics of penetration and reflection of ultrasonic waves, different reflected waves can be formed when the ultrasonic waves encounter different materials in the propagation process, and the sound velocity is different in the propagating media, so that a section of ultrasonic wave can be emitted to the chip by utilizing the characteristic, then the accurate material internal defect position is obtained through the calculation of a computer, and the ultrasonic scanning microscope can be used for the chip internal layered detection, this also solves the problems encountered with electron microscopes, which, with higher frequencies, the higher the precision of resolving defects, but the more complicated the technical difficulty, the scanning frequency of the ultrasonic scanning microscope in the chip industry is usually above 50MHz, the devices produced by domestic manufacturers are basically below 50MHz, and chip detection depends heavily on imported devices.
Disclosure of Invention
The purpose of the invention is as follows: in order to overcome the defects in the prior art, the application provides a testing method of a chip defect detection platform, and aims to solve the technical problems of how to quickly and accurately detect the height difference and the insufficient welding point of a welding spot caused by the gold wire on the surface of a chip in the production and solve the image capturing difficulty caused by insufficient depth of field of a camera, so that the appearance detection efficiency of the chip is improved.
The technical scheme is as follows: in order to achieve the purpose, the technical scheme adopted by the application is as follows:
a testing method of a chip defect detecting platform is applied to computer equipment, the computer equipment is used for controlling the chip defect detecting platform, the chip defect detection platform comprises an X-axis direction slide rail, a Y-axis direction slide rail, a Z-axis direction slide rail, a chip clamping block, a visual unit and a motion control card, the chip clamping block is used for placing a chip to be tested, the slide rail in the Y-axis direction drives the chip clamping block to move along a first direction, the X-axis direction slide rail drives the vision unit to move along a second direction, the Z-axis direction slide rail drives the vision unit to move along a third direction, the vision unit comprises a liquid lens and a camera, the camera is mounted on the liquid lens, the liquid lens is internally provided with a liquid lens control chip and a liquid lens, and the liquid lens control chip and the camera are in communication connection with the motion control card; the X-axis direction sliding rail is correspondingly connected with an X-axis motor, and the X-axis motor is correspondingly connected with an X-axis driver; the Y-axis direction sliding rail is correspondingly connected with a Y-axis motor, and the Y-axis motor is correspondingly connected with a Y-axis driver; the Z-axis direction sliding rail is correspondingly connected with a Z-axis motor, and the Z-axis motor is correspondingly connected with a Z-axis driver; the X-axis driver, the Y-axis driver and the Z-axis driver are in communication connection with the motion control card, and the motion control card is in communication connection with the computer equipment;
the method comprises the following steps:
s1, placing a chip to be tested on the chip clamping block;
s2, sending a first starting signal to a motion control card, wherein the motion control card drives an X-axis motor, a Y-axis motor and a Z-axis motor through an X-axis driver, a Y-axis driver and a Z-axis driver, and the X-axis motor, the Y-axis motor and the Z-axis motor drive an X-axis direction slide rail, a Y-axis direction slide rail and a Z-axis direction slide rail to move to corresponding X-axis preset coordinates, Y-axis preset coordinates and Z-axis preset coordinates;
s3, sending a second starting signal to the liquid lens control chip, wherein the second starting signal is used for controlling the liquid lens to automatically focus on the chip to be tested;
s4, controlling a camera to take a picture to obtain an initial image in response to the fact that the liquid lens finishes automatic focusing on the chip to be tested;
s5, changing the numerical value of the preset Z-axis coordinate, enabling the Z-axis direction slide rail to drive the liquid lens to move to the corresponding changed coordinate, and repeating the steps S3 and S4;
s6, repeating the step S5 for multiple times to obtain multiple groups of initial images, and carrying out image post-processing on the multiple groups of initial images to obtain processed images;
and S7, inputting the processed image into a deep learning model for analysis to obtain the defect detection result of the chip to be detected.
In a possible implementation manner, the step S2 further includes:
in response to the motion control card receiving the first starting signal, the motion control card generating a corresponding direction signal and a corresponding driving pulse signal and sending the direction signal and the driving pulse signal to the X-axis driver, the Y-axis driver and the Z-axis driver;
the X-axis driver, the Y-axis driver and the Z-axis driver generate corresponding pulses according to the direction signals and the driving pulse signals to drive the X-axis motor, the Y-axis motor and the Z-axis motor, and the X-axis motor, the Y-axis motor and the Z-axis motor drive the X-axis direction slide rail, the Y-axis direction slide rail and the Z-axis direction slide rail to move to the corresponding X-axis preset coordinate, Y-axis preset coordinate and Z-axis preset coordinate.
In a possible implementation manner, the step S3 further includes:
acquiring a focal length value of the liquid lens;
and translating the focal length value into a message and sending the message to the liquid lens control chip, wherein the liquid lens control chip controls the liquid lens to carry out automatic focusing.
In a possible implementation manner, the step S6 further includes:
extracting a plurality of groups of gold wire region images corresponding to the plurality of groups of initial images based on a Sobel algorithm;
and carrying out image synthesis on the multiple groups of gold thread area images to obtain a gold thread panoramic image.
In a possible implementation manner, the step S7 further includes:
performing Gaussian filtering on the gold wire panoramic image, calculating the gradient of the gold wire panoramic image in each direction based on a Sobel algorithm, and extracting a gold wire profile image on the surface of the chip to be detected by combining a Canny algorithm;
and inputting the gold wire outline image on the surface of the chip to be detected into a deep learning model for analysis to obtain a gold wire defect detection result on the surface of the chip to be detected.
In a possible implementation manner, the step S6 further includes:
extracting a plurality of groups of welding spot area images corresponding to a plurality of groups of initial images based on an SIFT algorithm;
and carrying out image synthesis on the multiple groups of welding spot area images to obtain a welding spot panoramic image.
In a possible implementation manner, the step S7 further includes:
preprocessing the welding spot panoramic image based on a GCN global contrast normalization algorithm to obtain a preprocessed welding spot panoramic image;
inputting the preprocessed welding spot panoramic image into a Relu convolution neural network model with a rectification linear unit for training to obtain a distinguishing welding spot model;
and obtaining a welding spot defect detection result on the surface of the chip to be detected based on the welding spot distinguishing model and a Canny edge extraction algorithm.
The beneficial effect that technical scheme that this application provided brought includes at least:
by controlling the chip defect detection platform to implement the test method through the computer equipment, the height difference and the solder joint rosin joint caused by gold wires on the surface of the chip in the production can be quickly and accurately detected, the image taking difficulty caused by insufficient depth of field of the camera is solved, and the appearance detection efficiency of the chip is improved.
Drawings
The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiment(s) of the application and together with the description serve to explain the application and not limit the application. In the drawings:
fig. 1 illustrates a schematic structural diagram of a chip defect inspection platform according to an exemplary embodiment of the present application;
fig. 2 is a schematic flowchart illustrating a testing method of a chip defect inspection platform according to an exemplary embodiment of the present application;
FIG. 3 is a flow chart illustrating another testing method for a chip defect inspection platform according to an exemplary embodiment of the present application;
fig. 4 is a flowchart illustrating another testing method for a chip defect inspection platform according to an exemplary embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments in the present application without making any creative effort belong to the protection scope of the present application.
The present application will be further described with reference to the following drawings and examples.
First, the terms referred to in the embodiments of the present application will be briefly described:
the SIFT feature point detection algorithm is an algorithm for detecting local features, and is used for obtaining features by solving the feature points in a picture and descriptors related to scale and orientation and carrying out image feature point matching. The SIFT feature detection mainly comprises the following 4 basic steps: (1) and (4) detecting an extreme value in a scale space, and searching image positions on all scales. Identifying potential interest points invariant to scale and rotation by a gaussian differential function; (2) and (4) positioning the key points, and determining the position and the scale by fitting a fine model at each candidate position. The selection of key points depends on their degree of stability; (3) and determining the direction, wherein one or more directions are allocated to each key point position based on the local gradient direction of the image. All subsequent operations on the image data are transformed with respect to critical directions, dimensions and locations, thereby providing invariance to these transformations; (4) keypoint description local gradients of an image are measured at a chosen scale in the neighborhood around each keypoint. These gradients are transformed into a representation that allows for relatively large local shape variations and illumination variations.
The RANSAC algorithm, which assumes that data includes correct data and abnormal data (or called noise), the correct data is marked as interior points (inerals), the abnormal data is marked as exterior points (outliers), and the RANSAC algorithm also assumes that a set of correct data is given, and a method capable of calculating model parameters conforming to the data exists; the core idea of the algorithm is randomness and hypothesis, wherein the randomness is to randomly select sampling data according to the occurrence probability of correct data, and the randomness simulation can approximately obtain correct results according to a law of large numbers; the hypothesis is that the sampled data are all correct data, then the correct data are used to calculate other points through the model satisfied by the problem, and then the result is scored.
The Sobel algorithm (namely the Sobel operator) is an important processing method in the field of computer vision, is mainly used for obtaining the first-order gradient of a digital image, and has the common application and physical significance of edge detection.
Gaussian filtering is linear smooth filtering, is suitable for eliminating Gaussian noise and is widely applied to the noise reduction process of image processing; generally speaking, gaussian filtering is a process of performing weighted average on the whole image, and the value of each pixel point is obtained by performing weighted average on the value of each pixel point and other pixel values in the neighborhood; the specific operation of gaussian filtering is: each pixel in the image is scanned using a template (or convolution, mask), and the weighted average gray value of the pixels in the neighborhood determined by the template is used to replace the value of the pixel in the center of the template.
The Canny algorithm, namely a Canny edge detection operator, is a multi-stage edge detection algorithm developed by john nf. Canny in 1986; more importantly, Canny has created a Computational theory of edge detection (Computational theory of edge detection) to explain how this technique works; the purpose of edge detection in general is to significantly reduce the data size of an image while preserving the original image attributes; there are many algorithms for edge detection, and although the Canny algorithm is a standard algorithm for edge detection, it is still widely used in research. The Canny edge detection algorithm can be divided into the following 5 steps: applying gaussian filtering to smooth the image with the aim of removing noise; finding intensity gradients (intensity gradients) of the image; applying a non-maximum suppression (non-maximum suppression) technique to eliminate edge false detection (which is not originally detected but detected); applying a dual threshold approach to determine possible (potential) boundaries; the boundaries are tracked using a hysteresis technique.
In deep learning, contrast generally refers to a standard deviation of pixels in an image or an image region, and contrast normalization includes global contrast normalization and local contrast normalization, which is a common data preprocessing method in deep learning, and is used to reduce variation in data, thereby reducing generalization errors and the size of a model required by fitting a training set.
Convolutional Neural Networks (CNN), which are a type of feed-forward Neural Networks (fed Neural Networks) containing convolution calculations and having deep structures, are one of the representative algorithms for deep learning (deep learning); convolutional Neural Networks have a representation learning (representation learning) capability, and can perform Shift-Invariant classification (Shift-Invariant classification) on input information according to a hierarchical structure thereof, and are also called Shift-Invariant Artificial Neural Networks (SIANN); the convolutional neural network is constructed by imitating a visual perception (visual perception) mechanism of a living being, can perform supervised learning and unsupervised learning, and has the advantages that convolutional kernel parameter sharing in an implicit layer and sparsity of interlayer connection, so that the convolutional neural network can learn grid-like topologic features such as pixels and audio with small calculation amount, has stable effect and has no additional feature engineering (feature engineering) requirement on data.
Relu, a Linear rectification function (Linear rectification function), also called a modified Linear unit, is an activation function (activation function) commonly used in artificial neural networks, and generally refers to a nonlinear function represented by a ramp function and a variation thereof; in a general sense, a linear rectification function refers to a ramp function in mathematics, i.e., (x) max (0, x).
Fig. 1 shows a schematic structural diagram of a chip defect detecting platform according to an exemplary embodiment of the present disclosure, where the chip defect detecting platform 1 includes a control cabinet base 11, a Y-axis direction slide rail 111 is installed on the control cabinet base 11, the Y-axis direction slide rail 111 is connected to a bearing table 113 through a first direction slide block 112 in a sliding manner, and a chip clamping block 17 is connected to the bearing table 113; the chip defect detection platform 1 further comprises a first support column 12, a second support column 13, a cross beam 14, a visual unit connecting seat 15 and a visual unit 16, wherein the first support column 12 and the second support column 13 are respectively connected with two sides of the rear end of the control cabinet base 11, the first support column 12 and the second support column 13 are symmetrically arranged, the first support column 12 and the second support column 13 are perpendicular to the control cabinet base 11, the top ends of the first support column 12 and the second support column 13 are respectively connected with two ends of the cross beam 14, and the cross beam 14 is perpendicular to the first support column 12 and the second support column 13; an X-axis direction slide rail 141 is installed on the front side of the cross beam 14, and the visual unit connecting seat 15 is connected with the X-axis direction slide rail 141 in a sliding manner through a second direction slide block 151; a Z-axis direction sliding rail 152 is installed on the front side of the visual unit connecting seat 15, the visual unit 16 is connected with the Z-axis direction sliding rail 152 in a sliding mode, the position of the visual unit 16 corresponds to that of the chip clamping block 17, and the visual unit 16 is in communication connection with the control cabinet base 11; the vision unit 16 includes a third direction slider 161, a vision unit mounting bracket 162, a liquid lens 163, and a camera 164; the third direction slider 161 is slidably connected to the Z-axis direction slide rail 152, the front side of the third direction slider 161 is connected to the visual unit mounting bracket 162, the visual unit mounting bracket 162 is perpendicular to the third direction slider 161, the liquid lens 163 is mounted in the visual unit mounting bracket 162, and the camera 164 is mounted on the liquid lens 163. The chip defect detection platform 1 further comprises a motion control card, a liquid lens control chip and a liquid lens are arranged in the liquid lens 163, and the liquid lens control chip and the camera 164 are in communication connection with the motion control card; the X-axis direction slide rail 141 is correspondingly connected with an X-axis motor, and the X-axis motor is correspondingly connected with an X-axis driver; the Y-axis direction slide rail 111 is correspondingly connected with a Y-axis motor, and the Y-axis motor is correspondingly connected with a Y-axis driver; the Z-axis direction slide rail 152 is correspondingly connected with a Z-axis motor, and the Z-axis motor is correspondingly connected with a Z-axis driver; the X-axis driver, the Y-axis driver and the Z-axis driver are in communication connection with a motion control card, and the motion control card is in communication connection with the computer equipment.
In this embodiment, the chip clamping block 17 is used for placing a chip to be tested, the Y-axis direction slide rail 111 drives the chip clamping block 17 to move along a first direction, the X-axis direction slide rail 141 drives the vision unit to move along a second direction, and the Z-axis direction slide rail 152 drives the vision unit 16 to move along a third direction.
In the embodiment of the present application, the liquid lens 163 has a model LM-23-09Y; a one-time telecentric lens is also arranged in the liquid lens 163, and the model of the one-time telecentric lens is TL1X110B 23; the model of the camera 164 is MV-CE120-10 GM; the camera 164 is further provided with a white coaxial light source, and the model of the white coaxial light source is TCL-6060-W.
Fig. 2 is a schematic flowchart illustrating a method for testing a chip defect inspection platform according to an exemplary embodiment of the present application, where the method is applied to a computer device, and the computer device is used to control the chip defect inspection platform, and the method includes:
s1, placing the chip to be tested on the chip clamping block;
s2, sending a first starting signal to a motion control card, wherein the motion control card drives an X-axis motor, a Y-axis motor and a Z-axis motor through an X-axis driver, a Y-axis driver and a Z-axis driver, and the X-axis motor, the Y-axis motor and the Z-axis motor drive an X-axis direction slide rail, a Y-axis direction slide rail and a Z-axis direction slide rail to move to corresponding X-axis preset coordinates, Y-axis preset coordinates and Z-axis preset coordinates;
s3, sending a second starting signal to the liquid lens control chip, wherein the second starting signal is used for controlling the liquid lens to automatically focus the chip to be tested;
s4, controlling the camera to shoot to obtain an initial image in response to the fact that the liquid lens finishes automatic focusing on the chip to be detected;
s5, changing the numerical value of the preset Z-axis coordinate, enabling the slide rail in the Z-axis direction to drive the liquid lens to move to the corresponding changed coordinate, and repeating the steps S3 and S4;
s6, repeating the step S5 for multiple times to obtain multiple groups of initial images, and carrying out image post-processing on the multiple groups of initial images to obtain processed images;
and S7, inputting the processed image into the deep learning model for analysis to obtain a defect detection result of the chip to be detected.
In the embodiment of the application, when the field of vision of the camera is not enough, the surface of the chip to be detected at the same height needs to be locally photographed in different regions for image capture, and then the local images are synthesized into a complete initial image based on an SIFT feature point detection algorithm in a spatial domain, and the method specifically comprises the following steps: extracting sift characteristic points; coarse matching of feature points is carried out by using a rapid nearest neighbor algorithm, and preliminary screening is carried out by using a threshold setting and a two-way cross inspection method; performing fine matching by using an RANSAC algorithm; image transformation, namely mapping different images to the same coordinate system, and calculating a transformation matrix according to the screened similar characteristic points; and (3) image fusion, namely transforming one image by using the calculated transformation matrix, overlapping the transformed image and the other image, and recalculating new pixel values in an overlapping area to finally obtain a complete initial image.
In the embodiment of the present application, the motor setting parameters corresponding to the X-axis motor, the Y-axis motor, and the Z-axis motor in step S4 at least include an axis coordinate, an axis type, an axis pulse equivalent, an axis start speed, an axis acceleration, an axis deceleration, and an axis emergency deceleration. The camera setting parameters at least include exposure time and setting data type.
Fig. 3 is a schematic flowchart illustrating another testing method for a chip defect inspection platform according to an exemplary embodiment of the present application, where the method is applied to a computer device, and the computer device is used to control the chip defect inspection platform, and the method includes:
s1, placing the chip to be tested on the chip clamping block;
s2, sending a first start signal to the motion control card, and in response to the motion control card receiving the first start signal, the motion control card generating a corresponding direction signal and a corresponding driving pulse signal and sending the direction signal and the driving pulse signal to the X-axis driver, the Y-axis driver and the Z-axis driver; the X-axis driver, the Y-axis driver and the Z-axis driver generate corresponding pulses according to the direction signals and the driving pulse signals to drive the X-axis motor, the Y-axis motor and the Z-axis motor, and the X-axis motor, the Y-axis motor and the Z-axis motor drive the X-axis direction slide rail, the Y-axis direction slide rail and the Z-axis direction slide rail to move to corresponding X-axis preset coordinates, Y-axis preset coordinates and Z-axis preset coordinates;
s3, acquiring a focal length value of the liquid lens; translating the focal length value into a message and sending the message to a liquid lens control chip, and controlling the liquid lens to carry out automatic focusing by the liquid lens control chip;
s4, controlling the camera to shoot to obtain an initial image in response to the fact that the liquid lens finishes automatic focusing on the chip to be detected;
s5, changing the numerical value of the preset coordinate of the Z axis, enabling the slide rail in the Z axis direction to drive the liquid lens to move to the corresponding changed coordinate, and repeating the steps S3 and S4;
s6, repeating the step S5 for multiple times to obtain multiple groups of initial images, and extracting multiple groups of gold wire area images corresponding to the multiple groups of initial images based on a Sobel algorithm; carrying out image synthesis on the multiple groups of gold thread area images to obtain a gold thread panoramic image;
s7, performing Gaussian filtering on the gold wire panoramic image, calculating the anisotropic gradient of the gold wire panoramic image based on a Sobel algorithm, and extracting a gold wire profile image on the surface of the chip to be detected by combining a Canny algorithm; and inputting the gold wire outline image of the surface of the chip to be detected into the deep learning model for analysis to obtain a gold wire defect detection result of the surface of the chip to be detected.
In this embodiment of the present application, after the step S3 translates the focal length value into a message and sends the message to the liquid lens control chip, the method further includes: the liquid lens control chip receives the message and analyzes the message to obtain Res; a PWM controller arranged in the liquid lens control chip generates a required waveform, and a full-bridge inverter arranged in the liquid lens control chip boosts the voltage and outputs the voltage to the liquid lens; the liquid lens receives the output voltage to generate a high-frequency electric field, and the space distribution of liquid in the liquid lens is changed, so that the diopter of the liquid lens is changed. Wherein, the relationship between Res and the output voltage is as follows: res 1023/46 (Vrms-24).
In the embodiment of the application, the testing method is implemented by controlling the chip defect detection platform through the computer equipment, so that the defect of different heights caused by gold wires on the surface of the chip in the production process can be quickly and accurately detected.
Fig. 4 is a schematic flowchart illustrating a testing method for a chip defect inspection platform according to another exemplary embodiment of the present application, where the method is applied to a computer device, and the computer device is used to control the chip defect inspection platform, and the method includes:
s1, placing the chip to be tested on the chip clamping block;
s2, sending a first start signal to the motion control card, and in response to the motion control card receiving the first start signal, the motion control card generating a corresponding direction signal and a corresponding driving pulse signal and sending the direction signal and the driving pulse signal to the X-axis driver, the Y-axis driver and the Z-axis driver; the X-axis driver, the Y-axis driver and the Z-axis driver generate corresponding pulses according to the direction signals and the driving pulse signals to drive the X-axis motor, the Y-axis motor and the Z-axis motor, and the X-axis motor, the Y-axis motor and the Z-axis motor drive the X-axis direction slide rail, the Y-axis direction slide rail and the Z-axis direction slide rail to move to corresponding X-axis preset coordinates, Y-axis preset coordinates and Z-axis preset coordinates;
s3, acquiring a focal length value of the liquid lens; translating the focal length value into a message and sending the message to a liquid lens control chip, and controlling the liquid lens to carry out automatic focusing by the liquid lens control chip;
s4, controlling the camera to shoot to obtain an initial image in response to the fact that the liquid lens finishes automatic focusing on the chip to be detected;
s5, changing the numerical value of the preset coordinate of the Z axis, enabling the slide rail in the Z axis direction to drive the liquid lens to move to the corresponding changed coordinate, and repeating the steps S3 and S4;
s6, repeating the step S5 for multiple times to obtain multiple groups of initial images, and extracting multiple groups of welding spot area images corresponding to the multiple groups of initial images based on the SIFT algorithm; carrying out image synthesis on the multiple groups of welding spot area images to obtain a welding spot panoramic image;
s7, preprocessing the welding spot panoramic image based on a GCN global contrast normalization algorithm to obtain a preprocessed welding spot panoramic image; inputting the preprocessed welding spot panoramic image into a Relu convolution neural network model with a rectification linear unit for training to obtain a distinguishing welding spot model; and obtaining a welding spot defect detection result of the surface of the chip to be detected based on the welding spot distinguishing model and combining with a Canny edge extraction algorithm.
In the embodiment of the application, the chip defect detection platform is controlled by the computer equipment to implement the test method, so that the defect of insufficient solder joint on the surface of the chip can be quickly and accurately detected.
In summary, the testing method is implemented by controlling the chip defect detection platform through the computer equipment, so that the height difference and the solder joint rosin joint caused by gold wires on the surface of the chip in production can be quickly and accurately detected, the image capturing difficulty caused by insufficient depth of field of the camera is solved, and the appearance detection efficiency of the chip is improved.
The above is only the preferred embodiment of the present application, and it should be noted that: it will be apparent to those skilled in the art that various modifications and enhancements can be made without departing from the principles of the application, and such modifications and enhancements are intended to be included within the scope of the application.
Claims (7)
1. A testing method of a chip defect detection platform is characterized in that the method is applied to computer equipment, the computer equipment is used for controlling the chip defect detection platform, the chip defect detection platform comprises an X-axis direction slide rail, a Y-axis direction slide rail, a Z-axis direction slide rail, a chip clamping block, a visual unit and a motion control card, the chip clamping block is used for placing a chip to be tested, the slide rail in the Y-axis direction drives the chip clamping block to move along a first direction, the X-axis direction slide rail drives the vision unit to move along a second direction, the Z-axis direction slide rail drives the vision unit to move along a third direction, the vision unit comprises a liquid lens and a camera, the camera is mounted on the liquid lens, the liquid lens is internally provided with a liquid lens control chip and a liquid lens, and the liquid lens control chip and the camera are in communication connection with the motion control card; the X-axis direction sliding rail is correspondingly connected with an X-axis motor, and the X-axis motor is correspondingly connected with an X-axis driver; the Y-axis direction sliding rail is correspondingly connected with a Y-axis motor, and the Y-axis motor is correspondingly connected with a Y-axis driver; the Z-axis direction sliding rail is correspondingly connected with a Z-axis motor, and the Z-axis motor is correspondingly connected with a Z-axis driver; the X-axis driver, the Y-axis driver and the Z-axis driver are in communication connection with the motion control card, and the motion control card is in communication connection with the computer equipment;
the method comprises the following steps:
s1, placing a chip to be tested on the chip clamping block;
s2, sending a first starting signal to a motion control card, wherein the motion control card drives an X-axis motor, a Y-axis motor and a Z-axis motor through an X-axis driver, a Y-axis driver and a Z-axis driver, and the X-axis motor, the Y-axis motor and the Z-axis motor drive an X-axis direction slide rail, a Y-axis direction slide rail and a Z-axis direction slide rail to move to corresponding X-axis preset coordinates, Y-axis preset coordinates and Z-axis preset coordinates;
s3, sending a second starting signal to the liquid lens control chip, wherein the second starting signal is used for controlling the liquid lens to automatically focus on the chip to be tested;
s4, controlling a camera to take a picture to obtain an initial image in response to the fact that the liquid lens finishes automatic focusing on the chip to be tested;
s5, changing the numerical value of the preset Z-axis coordinate, enabling the Z-axis direction slide rail to drive the liquid lens to move to the corresponding changed coordinate, and repeating the steps S3 and S4;
s6, repeating the step S5 for multiple times to obtain multiple groups of initial images, and carrying out image post-processing on the multiple groups of initial images to obtain processed images;
and S7, inputting the processed image into a deep learning model for analysis to obtain the defect detection result of the chip to be detected.
2. The method according to claim 1, wherein the step S2 further comprises:
responding to the motion control card receiving the first starting signal, the motion control card generating a corresponding direction signal and a corresponding driving pulse signal and sending the direction signal and the driving pulse signal to the X-axis driver, the Y-axis driver and the Z-axis driver;
the X-axis driver, the Y-axis driver and the Z-axis driver generate corresponding pulses according to the direction signals and the driving pulse signals to drive the X-axis motor, the Y-axis motor and the Z-axis motor, and the X-axis motor, the Y-axis motor and the Z-axis motor drive the X-axis direction slide rail, the Y-axis direction slide rail and the Z-axis direction slide rail to move to the corresponding X-axis preset coordinate, Y-axis preset coordinate and Z-axis preset coordinate.
3. The method according to claim 1, wherein the step S3 further comprises:
acquiring a focal length value of the liquid lens;
and translating the focal length value into a message and sending the message to the liquid lens control chip, wherein the liquid lens control chip controls the liquid lens to carry out automatic focusing.
4. The method according to claim 1, wherein the step S6 further comprises:
extracting a plurality of groups of gold wire area images corresponding to the plurality of groups of initial images based on a Sobel algorithm;
and carrying out image synthesis on the multiple groups of gold thread area images to obtain a gold thread panoramic image.
5. The method according to claim 4, wherein the step S7 further comprises:
performing Gaussian filtering on the gold wire panoramic image, calculating the gradient of the gold wire panoramic image in each direction based on a Sobel algorithm, and extracting a gold wire profile image on the surface of the chip to be detected by combining a Canny algorithm;
and inputting the gold wire outline image on the surface of the chip to be detected into a deep learning model for analysis to obtain a gold wire defect detection result on the surface of the chip to be detected.
6. The method according to claim 1, wherein the step S6 further comprises:
extracting a plurality of groups of welding spot area images corresponding to a plurality of groups of initial images based on an SIFT algorithm;
and carrying out image synthesis on the multiple groups of welding spot area images to obtain a welding spot panoramic image.
7. The method according to claim 6, wherein the step S7 further comprises:
preprocessing the welding spot panoramic image based on a GCN global contrast normalization algorithm to obtain a preprocessed welding spot panoramic image;
inputting the preprocessed welding spot panoramic image into a Relu convolution neural network model with a rectification linear unit for training to obtain a distinguishing welding spot model;
and obtaining a welding spot defect detection result of the surface of the chip to be detected based on the welding spot distinguishing model and a Canny edge extraction algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210521687.8A CN114965272A (en) | 2022-05-13 | 2022-05-13 | Testing method of chip defect detection platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210521687.8A CN114965272A (en) | 2022-05-13 | 2022-05-13 | Testing method of chip defect detection platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114965272A true CN114965272A (en) | 2022-08-30 |
Family
ID=82984114
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210521687.8A Withdrawn CN114965272A (en) | 2022-05-13 | 2022-05-13 | Testing method of chip defect detection platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114965272A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117074923A (en) * | 2023-10-11 | 2023-11-17 | 蓝芯存储技术(赣州)有限公司 | Chip burn-in test method, device, equipment and storage medium |
-
2022
- 2022-05-13 CN CN202210521687.8A patent/CN114965272A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117074923A (en) * | 2023-10-11 | 2023-11-17 | 蓝芯存储技术(赣州)有限公司 | Chip burn-in test method, device, equipment and storage medium |
| CN117074923B (en) * | 2023-10-11 | 2023-12-12 | 蓝芯存储技术(赣州)有限公司 | Chip burn-in test method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Automatic optical inspection system for IC molding surface | |
| CN111612737B (en) | Artificial board surface flaw detection device and detection method | |
| JP2017049974A (en) | Discriminator generator, quality determine method, and program | |
| US20110133054A1 (en) | Weighting surface fit points based on focus peak uncertainty | |
| JP4617970B2 (en) | Defect inspection apparatus and defect inspection method | |
| He et al. | A critical review for machining positioning based on computer vision | |
| CN110443791B (en) | Workpiece detection method and device based on deep learning network | |
| CN111932511B (en) | Electronic component quality detection method and system based on deep learning | |
| JPH07260701A (en) | Recognition method of area of inspection | |
| CN111982921A (en) | Hole defect detection method and device, conveying platform and storage medium | |
| CN108802051B (en) | System and method for detecting bubble and crease defects of linear circuit of flexible IC substrate | |
| CN111474179A (en) | Lens surface cleanliness detection device and method | |
| CN114965272A (en) | Testing method of chip defect detection platform | |
| CN1411046A (en) | Pattern evaluating apparatus, pattern evaluating method and program | |
| CN110738644A (en) | automobile coating surface defect detection method and system based on deep learning | |
| JP3009205B2 (en) | Inspection method and apparatus | |
| CN114092411A (en) | Efficient and rapid binocular 3D point cloud welding spot defect detection method | |
| CN113781415A (en) | Defect detection method, device, equipment and medium for X-ray image | |
| CN113705564A (en) | Pointer type instrument identification reading method | |
| CN113554054A (en) | Deep learning-based semiconductor chip gold wire defect classification method and system | |
| JP2008116206A (en) | Apparatus, method, and program for pattern size measurement | |
| JP2006292615A (en) | Visual examination apparatus, visual inspection method, program for making computer function as visual inspection apparatus, and recording medium | |
| Zou et al. | Laser-based precise measurement of tailor welded blanks: a case study | |
| TWI822126B (en) | Sample observation device, sample observation method and computer system | |
| KR102618194B1 (en) | Semiconductor package board processing hole 3d automatic process inspection method and system |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20220830 |