CN106248688B - A kind of wafer detection method for extracting signal based on FPGA - Google Patents

A kind of wafer detection method for extracting signal based on FPGA Download PDF

Info

Publication number
CN106248688B
CN106248688B CN201610765470.6A CN201610765470A CN106248688B CN 106248688 B CN106248688 B CN 106248688B CN 201610765470 A CN201610765470 A CN 201610765470A CN 106248688 B CN106248688 B CN 106248688B
Authority
CN
China
Prior art keywords
signal
data
window
frequency
wafer
Prior art date
Application number
CN201610765470.6A
Other languages
Chinese (zh)
Other versions
CN106248688A (en
Inventor
陈鲁
刘虹遥
马砚忠
张朝前
杨乐
路鑫超
Original Assignee
中国科学院嘉兴微电子仪器与设备工程中心
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 中国科学院嘉兴微电子仪器与设备工程中心 filed Critical 中国科学院嘉兴微电子仪器与设备工程中心
Priority to CN201610765470.6A priority Critical patent/CN106248688B/en
Publication of CN106248688A publication Critical patent/CN106248688A/en
Application granted granted Critical
Publication of CN106248688B publication Critical patent/CN106248688B/en

Links

Abstract

The present invention relates to wafer detection technique field, in particular to a kind of wafer based on FPGA detects method for extracting signal, comprising: acquisition wafer detects analog signal, and wafer detection analog signal is converted to digital signal;The data in digital signal are stored in buffer area according to timing, until data reach and specify number;Time Domain Processing and frequency domain processing are carried out to the data of buffer area;Time-domain signal and frequency-region signal are searched, is uploaded to host computer after time-domain signal and frequency-region signal are merged judgement.Wafer provided by the invention based on FPGA detects method for extracting signal, improves the extraction efficiency of the analog signal of wafer detecting apparatus detection, more stable to the data processing in the wafer signal of extraction, accurate and sensitive.

Description

A kind of wafer detection method for extracting signal based on FPGA

Technical field

The present invention relates to wafer detection technique field, in particular to a kind of wafer based on FPGA detects signal extraction side Method.

Background technique

Wafer growth is the basis of semiconductor industry, due to being influenced by factors such as material, equipment and environment, in crystalline substance Circular surfaces inevitably will appear particle contamination, therefore how be protected by indexs such as quantity, the density of detection particle contamination The normal growth of card wafer has become key link.Have using the scan-type detection method of light scattering quickly, without additional dirt The advantages such as dye are the main methods that current semicon industry carries out wafer complete detection.Its basic principle is to utilize small light spot Oblique incidence exposes to crystal column surface, and since crystal column surface is very smooth, when not polluting at hot spot irradiation, incident light will be whole It is reflected away with equal angular from the other side;It will be scattered when there is pollution, scattering light is extended from surrounding, and scattered light intensity takes Certainly in the size of pollution and refractive index, by successively scanning whole wafer surface using laser, it can realize that wafer detects.Due to When pollution size is smaller on wafer, scattering light is very faint, therefore how to improve the significant challenge that detection sensitivity is the method. The cardinal principle of the wafer detection method and wafer detecting apparatus that occur at present is the hot spot using synthesis light intensity period profile The periodic modulation for realizing scattered signal, is concentrated to signal at some specific frequency, since system noise is close uniform in wide range Distribution analyzes the influence for eliminating partial noise by carrying out signal near specific frequency to improve sensitivity, but for crystalline substance The analog signal of loop truss device detection is extracted and processing method not yet occurs.

Summary of the invention

The present invention improves wafer detecting apparatus inspection by providing a kind of wafer detection method for extracting signal based on FPGA The extraction efficiency of the analog signal of survey, it is more stable, sensitive to the processing of the wafer signal of extraction, it ensure that the standard of data processing True property.

The present invention provides a kind of, and the wafer based on FPGA detects method for extracting signal, comprising:

It acquires wafer and detects analog signal, wafer detection analog signal is converted into digital signal;

The data in the digital signal are stored in buffer area according to timing;

The data for taking the buffer area carry out at Time Domain Processing and frequency domain the data of the buffer area Reason;

Time-domain signal and frequency-region signal are searched, is uploaded to after the time-domain signal and the frequency-region signal are merged judgement Position machine.

It is further, described that the data in the digital signal are stored in buffer area according to timing, comprising:

According to timing by the data pick-up in the digital signal be at least two long fragments;

The different segments is stored respectively in different data groups, and the data group is stored in described delay Rush area.

Further, the data for taking the buffer area, comprising:

Window look-up table controls the count value of time segment counter, according to the data group of timing starting different labels;

Data in the data group of enabling are subjected to time domain and frequency domain processing.

It is further, described that the data in the digital signal are stored in buffer area according to timing, further includes:

It is arranged and the window of the data group corresponding number;

The loop parameter of the segment is stored in the window;

Segment in the data group searches the corresponding window by window look-up table, and obtains corresponding segment Loop parameter.

Further, the Time Domain Processing, comprising:

The number of data group where marking the data;

The loop parameter in the corresponding window of data group is obtained by window look-up table;

Find the maximum value of the data in the data group of place;

By the maximum value of the data and given time domain threshold value comparison, if more than the time domain threshold value, then by described in most Big value time-domain signal of window as where temporarily stores.

Further, the frequency domain processing, comprising:

The window number of data group where marking the data;

The loop parameter in the corresponding window of data group is obtained by window look-up table;

Fourier transformation is carried out to the data and obtains frequency domain distribution;

It is if more than the frequency domain threshold value, then frequency range is strong by the maximum value of frequency range intensity and given frequency domain threshold value comparison The maximum value of degree frequency-region signal of window as where temporarily stores.

It is further, described to be uploaded to host computer after the time-domain signal and frequency-region signal merging judgement, comprising:

When the time-domain signal and the frequency-region signal exist simultaneously, then determine at the corresponding wafer position of current window There are large-size particles, while the time-domain signal being stored and is uploaded to the host computer;

When there is only the frequency-region signal, then determine at the corresponding wafer position of current window there are small sized particles, The frequency-region signal is stored and is uploaded to the host computer simultaneously;

In the absence of the time-domain signal and frequency-region signal are equal, then determine not deposit at the corresponding wafer position of current window In particle;

When there is only the time-domain signal, then determine testing result exception, by the time-domain signal data store and Mistake memory block marks current window number.

Further, further includes:

Clock signal is introduced in the digital signal, for carrying out clock control.

Further, when acquisition wafer detection analog signal, the unit acquisition capacity of the analog signal is 14bit, is adopted Integrate frequency as 500MHz.

Further, the window look-up table and the window are stored in DDR3.

One or more technical solutions provided by the invention, at least have following beneficial effect or advantage:

Wafer provided in an embodiment of the present invention detection method for extracting signal and device, changing this method will will be digital according to timing Data in signal are stored in buffer area;The data of buffer area are sent into fpga chip and carry out time domain and frequency domain processing;Search time domain Signal and frequency-region signal are uploaded to host computer after time-domain signal and frequency-region signal are merged judgement.According to timing by digital signal In data be stored in buffer area, then from buffer area take data carry out time domain and frequency domain processing, improve wafer detecting apparatus inspection The extraction efficiency of the analog signal of survey, it is more stable, accurate to the data processing in the wafer signal of extraction.

Wafer detection method for extracting signal and device provided in an embodiment of the present invention, are provided with for storing loop parameter Multiple windows, each window is relatively independent, convenient for modifying different loop parameters without influencing other windows, improves The treatment effeciency of data.

Detailed description of the invention

Fig. 1 is that wafer provided in an embodiment of the present invention detects method for extracting signal flow chart.

Specific embodiment

The embodiment of the present invention improves wafer detection by providing a kind of wafer detection method for extracting signal based on FPGA The extraction efficiency of the analog signal of device detection, it is more stable, sensitive to the processing of the wafer signal of extraction, it ensure that at data The accuracy of reason.

Referring to Fig. 1, the embodiment of the invention provides a kind of, and the wafer based on FPGA detects method for extracting signal, comprising:

Step S1, acquisition wafer detects analog signal, and wafer detection analog signal is converted to digital signal.Step S1 tool Body includes: step S11, from wafer detecting apparatus acquisition wafer detection analog signal, and wafer detection analog signal is converted to number Word signal.When acquiring wafer detection analog signal, the unit acquisition capacity of analog signal is 14bit, frequency acquisition 500MHz.

Step S2, the data in digital signal are stored in buffer area according to timing, until data reach and specify number.Step S2 is specifically included: being at least two long fragments by the data pick-up in digital signal step S21, according to timing.Step S22, Different segments is stored respectively in different data groups, marks the number of each data group, and data group is stored in Buffer area.Step S23, the window of setting and data group corresponding number;The loop parameter of stored fragments in window;In data group Segment corresponding window is searched by window look-up table, and obtain the loop parameter of corresponding segment.Wherein, window look-up table And window is stored in DDR3, when window look-up table defines total window number, window number, length of window (i.e. FFT length), window Between section (defining when starting-window), time domain threshold value, frequency dividing when frequency domain window mouth position.Step S24, whether detection data Reach specified number.Step S25, when detect data reach specify number when, stop to buffer area be stored in data.

Step S3, the data for taking buffer area carry out Time Domain Processing to the data of buffer area and frequency domain are handled.Step S3 tool Body includes: step S31, window look-up table by the count value of FIFO real-time control time segment counter, is started not according to timing With the data group of label.Step S32, it is corresponding to obtain data group by window look-up table for the number of data group where flag data Window in loop parameter, where finding in data group data maximum value;By the maximum value of data and given time domain threshold Value compares, and if more than time domain threshold value, then using maximum value, the time-domain signal of window as where is temporarily stored.Step 33, reference numerals According to the number of place data group, the loop parameter in the corresponding window of data group is obtained by window look-up table, data are carried out Fourier transformation obtains frequency domain distribution;By the maximum value of frequency range intensity and given frequency domain threshold value comparison, if more than frequency domain threshold value, Then using the maximum value of frequency range intensity, the frequency-region signal of window as where is temporarily stored.

Step S4, time-domain signal and frequency-region signal are searched, is uploaded to after time-domain signal and frequency-region signal are merged judgement Position machine.Step S4 specifically includes step S41, searches time-domain signal and frequency-region signal.Step S42, when time-domain signal and frequency domain are believed When number existing simultaneously, then determine that there are large-size particles at the corresponding wafer position of current window, while time-domain signal being stored And it is uploaded to host computer;When there is only frequency-region signal, then determine that there are small sizes at the corresponding wafer position of current window Grain, while frequency-region signal being stored and is uploaded to host computer;In the absence of time-domain signal and frequency-region signal are equal, then determine current Particle is not present at the corresponding wafer position of window;When there is only time-domain signal, then determines testing result exception, time domain is believed Number stores and marks current data set number in wrong memory block.

Step S5, clock signal is introduced in a digital signal, for carrying out clock control.

The wafer detection method for extracting signal provided by the invention based on FPGA is carried out below with reference to specific embodiment Illustrate:

In the present embodiment, the scanning area of wafer is divided into 50 regions, 50 scanning areas crystalline substance different for 50 Round wires speed, the wafer detection method for extracting signal provided in this embodiment based on FPGA include:

Analog signal is detected from wafer detecting apparatus acquisition wafer, wafer detection analog signal is converted into digital signal. When acquiring wafer detection analog signal, the unit acquisition capacity of analog signal is 14bit, frequency acquisition 500MHz.

According to timing by the data pick-up in digital signal be 50 long fragments.

Different segments being stored respectively in different data groups, 50 long fragments respectively correspond 50 data groups, 50 data components do not carry out label;In timing arrangement, the rear end of the data and previous window of the front end 1/2 of latter window Data duplication;50 data groups are stored in buffer area simultaneously.

Setting 50 and the one-to-one window of data group;The circulation ginseng of long fragment is stored in 50 windows respectively Number;Long fragment in data group searches corresponding window by window look-up table, and obtains the loop parameter of corresponding segment. Wherein, window look-up table and window are stored in DDR3.

Whether detection data reaches 50 specified long fragments.

When detecting that data reach 50 specified long fragments, stop being stored in data to buffer area.

Window look-up table passes through the count value of FIFO real-time control time segment counter, according to timing starting different labels Data group, window look-up table are worked by FIFO real-time control other function module.

The number of data group where flag data obtains the circulation in the corresponding window of data group by window look-up table and joins Number (loop parameter is parameter needed for Time Domain Processing), finds the maximum value of data in the window of place;By the maximum value of data With given time domain threshold value comparison, if more than time domain threshold value, then using maximum value, the time-domain signal of window as where is temporarily stored.

The number of data group where flag data obtains the circulation in the corresponding window of data group by window look-up table and joins Data are carried out Fourier transformation and obtain frequency domain distribution by number (loop parameter is parameter needed for frequency domain processing);Frequency range is strong The maximum value of degree and given frequency domain threshold value comparison, if more than time domain threshold value, then using the maximum value of frequency range intensity as place window The frequency-region signal of mouth temporarily stores.

From scratchpad area (SPA) domain lookup time-domain signal and frequency-region signal.

When time-domain signal and frequency-region signal exist simultaneously, then determine that there are big rulers at the corresponding wafer position of current window Very little particle, while time-domain signal being stored and is uploaded to host computer;When there is only frequency-region signal, then determine that current window is corresponding Wafer position at there are small sized particles conditions, while frequency-region signal being stored and is uploaded to host computer;When time-domain signal and frequency In the absence of domain signal is equal, then determine that there is no particles at the corresponding wafer position of current window;When there is only time-domain signal, Then determine testing result exception, time-domain signal data are stored and mark current window number in wrong memory block.

Clock signal is introduced in a digital signal, for carrying out clock control.

The embodiment of the invention also provides a kind of, and the wafer based on FPGA detects signal extracting device, comprising:

Wafer detection analog signal is converted to digital letter for acquiring wafer detection analog signal by signal acquisition unit Number.Wherein, signal acquisition unit includes: collection of simulant signal module and analog-to-digital conversion module.Collection of simulant signal module is used for Analog signal is detected from wafer detecting apparatus acquisition wafer.Analog-to-digital conversion module is used to wafer detection analog signal being converted to number Word signal.When acquiring wafer detection analog signal, the unit acquisition capacity of analog signal is 14bit, frequency acquisition 500MHz.

Data in digital signal are stored in buffer area according to timing, until data reach specified number by data pick-up unit Mesh.Wherein, data pick-up unit includes: time series data abstraction module, memory module, window setting module and detection module.When Sequence data extraction module is used to according to timing be at least two long fragments by the data pick-up in digital signal.Memory module is used In different segments to be stored respectively in different data groups, the window number of each data group is marked, and data group is deposited It is stored in buffer area.Window setting module is for being arranged and the window of data group corresponding number;The circulation of stored fragments in window Parameter;Segment in data group searches corresponding window by window look-up table, and obtains the loop parameter of corresponding segment.Its In, window look-up table and window are stored in DDR3, and window look-up table defines total window number, window number, length of window (FFT Length), window periods (defining when starting-window), time domain threshold value, frequency dividing when frequency domain window mouth position.Detection module is used Whether reach specified number in detection data;When detect data reach specify number when, stop to buffer area be stored in data.

Searching unit takes the data of buffer area, carries out Time Domain Processing to the data of buffer area and frequency domain is handled.Wherein, Searching unit includes: window starting module, and window look-up table passes through the count value of FIFO real-time control time segment counter, according to The data group of timing starting different labels;And the data in the data group of enabling are sent into time domain processing module and frequency domain processing Module carries out Time Domain Processing and frequency domain processing.Number of the time domain processing module for data group where flag data, passes through window Look-up table obtains the loop parameter (loop parameter is parameter needed for Time Domain Processing) in the corresponding window of data group, finds institute The maximum value of data in window;It then will most if more than time domain threshold value by the maximum value of data and given time domain threshold value comparison Big value time-domain signal of window as where temporarily stores.Number of the frequency domain processing module for data group where flag data, Obtaining the loop parameter in the corresponding window of data group by window look-up table, (loop parameter is ginseng needed for frequency domain processing Number), Fourier transformation is carried out to data and obtains frequency domain distribution;By the maximum value of frequency range intensity and given frequency domain threshold value comparison, If more than time domain threshold value, then using the maximum value of frequency range intensity, the frequency-region signal of window as where is temporarily stored.

Judging unit searches time-domain signal and frequency-region signal, is uploaded to after time-domain signal and frequency-region signal are merged judgement Host computer.Wherein, judging unit includes: signal searching module and signal judgment module.Signal searching module is for searching time domain Signal and frequency-region signal.Signal judgment module is used to then determine current window when time-domain signal is existed simultaneously with frequency-region signal There are large-size particles at corresponding wafer position, while time-domain signal being stored and is uploaded to host computer;When there is only frequency domains When signal, then determine that there are small sized particles conditions at the corresponding wafer position of current window, at the same by frequency-region signal storage and on Reach host computer;In the absence of time-domain signal and frequency-region signal are equal, then determine not deposit at the corresponding wafer position of current window In particle;When there is only time-domain signal, then determines testing result exception, time-domain signal data are stored and in wrong memory block Mark current window number.

Clock unit introduces clock signal in a digital signal, for carrying out clock control.

One or more technical solutions provided in an embodiment of the present invention, at least have it is following the utility model has the advantages that

Wafer provided in an embodiment of the present invention detection method for extracting signal and device, changing this method will will be digital according to timing Data in signal are stored in buffer area;The data of buffer area are sent into fpga chip and carry out time domain and frequency domain processing;Search time domain Signal and frequency-region signal are uploaded to host computer after time-domain signal and frequency-region signal are merged judgement.According to timing by digital signal In data be stored in buffer area, then from buffer area take data carry out time domain and frequency domain processing, improve wafer detecting apparatus inspection The extraction efficiency of the analog signal of survey, it is more stable, accurate to the data processing in the wafer signal of extraction.

Wafer detection method for extracting signal and device provided in an embodiment of the present invention, are provided with for storing loop parameter Multiple windows, each window is relatively independent, convenient for modifying different loop parameters without influencing other windows, improves The treatment effeciency of data.

Wafer detection method for extracting signal and device provided in an embodiment of the present invention, collect every one piece of data and distribute incessantly Carry out signal extraction to window, in addition to the window for coming head end, each window front end be arranged a certain proportion of data with Previous window Back end data is identical, improves the sensitivity of Data Detection.

It should be noted last that the above specific embodiment is only used to illustrate the technical scheme of the present invention and not to limit it, Although being described the invention in detail referring to example, those skilled in the art should understand that, it can be to the present invention Technical solution be modified or replaced equivalently, without departing from the spirit and scope of the technical solution of the present invention, should all cover In the scope of the claims of the present invention.

Claims (8)

1. a kind of wafer based on FPGA detects method for extracting signal characterized by comprising
It acquires wafer and detects analog signal, wafer detection analog signal is converted into digital signal;
The data in the digital signal are stored in buffer area according to timing;
The data for taking the buffer area carry out Time Domain Processing to the data of the buffer area and frequency domain are handled;
Time-domain signal and frequency-region signal are searched, it is upper by being uploaded to after the time-domain signal and frequency-region signal merging judgement Machine;
It is wherein, described that the data in the digital signal are stored in buffer area according to timing, comprising:
According to timing by the data pick-up in the digital signal be at least two long fragments;
The different segments is stored respectively in different data groups, and the data group is stored in the buffering Area;
It is arranged and the window of the data group corresponding number;
The loop parameter of the segment is stored in the window, loop parameter is ginseng needed for Time Domain Processing and frequency domain processing Number;
Segment in the data group searches the corresponding window by window look-up table, and obtains the circulation of corresponding segment Parameter.
2. wafer based on FPGA detects method for extracting signal as described in claim 1, which is characterized in that it is described take it is described The data of buffer area, comprising:
Window look-up table controls the count value of time segment counter, according to the data group of timing starting different labels;
Data in the data group of enabling are subjected to time domain and frequency domain processing.
3. the wafer based on FPGA detects method for extracting signal as claimed in claim 2, which is characterized in that at the time domain Reason, comprising:
The number of data group where marking the data;
The loop parameter in the corresponding window of data group is obtained by window look-up table;
Find the maximum value of the data in the data group of place;
By the maximum value of the data and given time domain threshold value comparison, if more than the time domain threshold value, then by the maximum value Time-domain signal as place window temporarily stores.
4. the wafer as described in any one of claims 1-3 based on FPGA detects method for extracting signal, which is characterized in that described Frequency domain processing, comprising:
The window number of data group where marking the data;
The loop parameter in the corresponding window of data group is obtained by window look-up table;
Fourier transformation is carried out to the data and obtains frequency domain distribution;
By the maximum value of frequency range intensity and given frequency domain threshold value comparison, if more than the frequency domain threshold value, then by frequency range intensity Maximum value frequency-region signal of window as where temporarily stores.
5. the wafer as described in any one of claims 1-3 based on FPGA detects method for extracting signal, which is characterized in that described Host computer is uploaded to after the time-domain signal and the frequency-region signal are merged judgement, comprising:
When the time-domain signal and the frequency-region signal exist simultaneously, then determine exist at the corresponding wafer position of current window Large-size particle, while the time-domain signal being stored and is uploaded to the host computer;
When there is only the frequency-region signal, then determine that there are small sized particles at the corresponding wafer position of current window, simultaneously The frequency-region signal is stored and is uploaded to the host computer;
In the absence of the time-domain signal and frequency-region signal are equal, then determine at the corresponding wafer position of current window there is no Grain;
When there is only the time-domain signal, then determines testing result exception, the time-domain signal data are stored and in mistake Memory block marks current window number.
6. the wafer based on FPGA detects method for extracting signal as claimed in claim 5, which is characterized in that further include:
Clock signal is introduced in the digital signal, for carrying out clock control.
7. the wafer based on FPGA detects method for extracting signal as claimed in claim 6, which is characterized in that acquisition wafer detection When analog signal, the unit acquisition capacity of the analog signal is 14bit, frequency acquisition 500MHz.
8. the wafer based on FPGA detects method for extracting signal as described in claim 1, which is characterized in that the window is searched Table and the window are stored in DDR3.
CN201610765470.6A 2016-08-30 2016-08-30 A kind of wafer detection method for extracting signal based on FPGA CN106248688B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610765470.6A CN106248688B (en) 2016-08-30 2016-08-30 A kind of wafer detection method for extracting signal based on FPGA

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610765470.6A CN106248688B (en) 2016-08-30 2016-08-30 A kind of wafer detection method for extracting signal based on FPGA

Publications (2)

Publication Number Publication Date
CN106248688A CN106248688A (en) 2016-12-21
CN106248688B true CN106248688B (en) 2019-04-16

Family

ID=57596858

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610765470.6A CN106248688B (en) 2016-08-30 2016-08-30 A kind of wafer detection method for extracting signal based on FPGA

Country Status (1)

Country Link
CN (1) CN106248688B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1092890A (en) * 1997-09-01 1998-04-10 Shin Etsu Handotai Co Ltd Method of measuring number of particles generated from semiconductor wafer
CN102023168A (en) * 2010-11-08 2011-04-20 北京大学深圳研究生院 Method and system for detecting chips on semiconductor wafer surface
CN102759533A (en) * 2011-04-27 2012-10-31 中国科学院微电子研究所 Wafer detecting method and wafer detecting device
CN103018258A (en) * 2011-09-23 2013-04-03 中国科学院微电子研究所 Wafer detection method and wafer detection apparatus
CN103748454A (en) * 2011-07-12 2014-04-23 科磊股份有限公司 Wafer inspection
JP2015008062A (en) * 2013-06-25 2015-01-15 株式会社日立ハイテクノロジーズ Charged particle beam device and sample inspection method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5319876B2 (en) * 2006-07-31 2013-10-16 株式会社日立ハイテクノロジーズ Surface inspection apparatus and surface inspection method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1092890A (en) * 1997-09-01 1998-04-10 Shin Etsu Handotai Co Ltd Method of measuring number of particles generated from semiconductor wafer
CN102023168A (en) * 2010-11-08 2011-04-20 北京大学深圳研究生院 Method and system for detecting chips on semiconductor wafer surface
CN102759533A (en) * 2011-04-27 2012-10-31 中国科学院微电子研究所 Wafer detecting method and wafer detecting device
CN103748454A (en) * 2011-07-12 2014-04-23 科磊股份有限公司 Wafer inspection
CN103018258A (en) * 2011-09-23 2013-04-03 中国科学院微电子研究所 Wafer detection method and wafer detection apparatus
JP2015008062A (en) * 2013-06-25 2015-01-15 株式会社日立ハイテクノロジーズ Charged particle beam device and sample inspection method

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"Detection of defects on the surface of a semiconductor by terahertz surface plasmon polaritons;TAO YANG et al.;《Applied Optics》;20160520;第55卷(第15期);摘要、第5-6节、图4-5
一种基于光散射原理在线测量晶圆表面质量的方法;张楠 等;《传感器与微系统》;20071231;第26卷(第8期);第104-106页

Also Published As

Publication number Publication date
CN106248688A (en) 2016-12-21

Similar Documents

Publication Publication Date Title
CN104005325B (en) Based on the depth and grayscale images pavement crack detection apparatus and method
CN101672839B (en) Device and method for detecting hatching egg incubation quality based on computer vision
Parrish et al. Empirical comparison of full-waveform lidar algorithms
Cotton et al. The effective density of small ice particles obtained from in situ aircraft observations of mid‐latitude cirrus
CN103235830A (en) Unmanned aerial vehicle (UAV)-based electric power line patrol method and device and UAV
US6687396B1 (en) Optical member inspection apparatus, image-processing apparatus, image-processing method, and computer readable medium
CN104198482A (en) Test paper reading method and test paper reading device
CN103870845B (en) Novel K value optimization method in point cloud clustering denoising process
Álvarez et al. Improvement of plankton biovolume estimates derived from image-based automatic sampling devices: application to FlowCAM
Lin et al. Tree species classification based on explicit tree structure feature parameters derived from static terrestrial laser scanning data
Tsai et al. Multiscale crack fundamental element model for real-world pavement crack classification
JP2011523706A (en) Signal search with 3D bitmap
CN101726693B (en) Method for seeking discharge regions of power devices on ultraviolet images
CN104021574B (en) Pavement disease automatic identifying method
Ates et al. An image-processing based automated bacteria colony counter
Chen et al. Automated inspection of engineering ceramic grinding surface damage based on image recognition
CN102435924A (en) Automatic positioning method and system for cable local discharge single end of OWTS (Oscillating Wave Test System) detection device
CN102968799A (en) Integral image-based quick ACCA-CFAR SAR (Automatic Censored Cell Averaging-Constant False Alarm Rate Synthetic Aperture Radar) image target detection method
US20070041422A1 (en) Automated binary processing of thermographic sequence data
CN103163141A (en) Strip steel surface on-line inspection system and method based on embedded image processing system
US8449176B2 (en) Automated binary processing of thermographic sequence data
Barchyn et al. Field comparison of four piezoelectric sensors for detecting aeolian sediment transport
Luo et al. Automatic positioning for picking point of grape picking robot in natural environment
CN101275824A (en) Method for detecting rice granule type
JP2019074537A (en) Methods for detecting coincident sample events

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
GR01 Patent grant