CN112596965B - Digital image bad cluster statistical method and integrated circuit automatic tester - Google Patents
Digital image bad cluster statistical method and integrated circuit automatic tester Download PDFInfo
- Publication number
- CN112596965B CN112596965B CN202011479362.5A CN202011479362A CN112596965B CN 112596965 B CN112596965 B CN 112596965B CN 202011479362 A CN202011479362 A CN 202011479362A CN 112596965 B CN112596965 B CN 112596965B
- Authority
- CN
- China
- Prior art keywords
- bad cluster
- format
- image data
- digital image
- preset
- 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.)
- Active
Links
- 238000007619 statistical method Methods 0.000 title claims abstract description 33
- 239000011159 matrix material Substances 0.000 claims abstract description 71
- 238000004364 calculation method Methods 0.000 claims abstract description 54
- 102000008297 Nuclear Matrix-Associated Proteins Human genes 0.000 claims abstract description 12
- 108010035916 Nuclear Matrix-Associated Proteins Proteins 0.000 claims abstract description 12
- 210000000299 nuclear matrix Anatomy 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims description 24
- 230000002159 abnormal effect Effects 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 9
- 238000009499 grossing Methods 0.000 claims description 8
- 238000000034 method Methods 0.000 abstract description 18
- 238000010998 test method Methods 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 17
- 230000008569 process Effects 0.000 description 9
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000001502 supplementing effect Effects 0.000 description 2
- 238000012896 Statistical algorithm Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/22—Detection or location of defective computer hardware by testing during standby operation or during idle time, e.g. start-up testing
- G06F11/2205—Detection or location of defective computer hardware by testing during standby operation or during idle time, e.g. start-up testing using arrangements specific to the hardware being tested
- G06F11/2215—Detection or location of defective computer hardware by testing during standby operation or during idle time, e.g. start-up testing using arrangements specific to the hardware being tested to test error correction or detection circuits
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Data Mining & Analysis (AREA)
- Computational Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
Abstract
The invention provides a digital image bad cluster statistical method, a CIS chip test method and an integrated circuit automatic tester. The digital image bad cluster statistical method comprises the following steps: decomposing the image data in the first format into a plurality of submatrices with preset sizes, and carrying out preset calculation on each submatrix and a nuclear matrix with preset sizes; wherein, the elements in the kernel matrix conform to the following rules: the result of multiplication of any two combinations comprising at least two elements is different. The configuration ensures that the number of the bad cluster type images with the preset size can be counted only by carrying out one-time scanning treatment and calculation on the original digital image, has the characteristics of rigorousness and high efficiency, and solves the problems of inaccurate counting result and low counting efficiency caused by the fact that the counting method of the bad clusters of the images in the prior art is not strict.
Description
Technical Field
The invention relates to the field of semiconductor testing, in particular to a digital image bad cluster statistical method, a CIS chip testing method and an integrated circuit automatic testing machine.
Background
The current domestic CIS chip test evaluation bad cluster statistics or digital image evaluation bad cluster statistics generally have the following defects:
1, the defect statistical algorithm is not strict enough, and the statistical result form is inconvenient to process;
2. to facilitate statistics and computation, single statistics block regions are limited, such as 3*3;
3. different types of bad clusters (defect combinations) are not counted enough or repeated, and multiple types of operators and multiple scans and calculations are required.
In a word, in the prior art, the statistics method of the bad clusters of the images is not strict enough, so that the statistics result is not accurate enough and the statistics efficiency is low.
Disclosure of Invention
The invention aims to provide a digital image bad cluster statistical method, a CIS chip test method and an integrated circuit automatic test machine, which are used for solving the problems of inaccurate statistical results and low statistical efficiency caused by the fact that the statistical method of the image bad clusters is not strict in the prior art.
In order to solve the above technical problem, according to a first aspect of the present invention, there is provided a digital image bad cluster statistical method, which is characterized in that the digital image bad cluster statistical method includes: decomposing the image data in the first format into a plurality of submatrices with preset sizes, and carrying out preset calculation on each submatrix and a nuclear matrix with preset sizes; wherein, the elements in the kernel matrix conform to the following rules: the result of multiplication of any two combinations comprising at least two elements is different.
Optionally, the kernel matrix is filled with different prime numbers.
Optionally, the kernel matrix is filled with a minimum of N prime numbers, where N is equal to the product of the number of rows of the preset size and the number of columns of the preset size.
Optionally, the first format is that 0 or 1 is used for assigning a pixel of the image data, wherein 0 is used for assigning a normal pixel, and 1 is used for assigning an abnormal pixel; and meanwhile, the elements in the submatrices with equal row numbers and equal column numbers and the elements in the core matrix have corresponding relations, and the preset calculation comprises the steps of multiplying the elements in the core matrix corresponding to the elements with the value of 1 in the submatrices.
Optionally, the digital image bad cluster statistical method includes: setting a threshold value, performing assignment operation on image data in a second format, assigning a point with an absolute value greater than or equal to the threshold value in the image data in the second format to be 1, and assigning the rest pixel points to be 0 to obtain the image data in the first format.
Optionally, the digital image bad cluster statistical method includes: and carrying out smoothing processing on the image data in the original format to obtain the image data in the smoothed format, and subtracting the value of each pixel point in the image data in the original format from the value of the corresponding pixel point in the image data in the smoothed format to obtain the image data in the second format.
Optionally, the digital image bad cluster statistical method includes: obtaining a bad cluster distribution set, wherein the bad cluster distribution set comprises at least one bad cluster type image with the preset size in the first format; and carrying out the preset calculation on each bad cluster type image and the core matrix in the bad cluster distribution set to obtain a bad cluster type calculation result, and generating a lookup table between the corresponding bad cluster type image and the calculation result.
Optionally, the digital image bad cluster statistical method includes: and obtaining a calculation result of the submatrix and the nuclear matrix after the preset calculation, and comparing the calculation result with the lookup table to obtain bad cluster statistical information.
In order to solve the above technical problem, according to a second aspect of the present invention, there is provided a CIS chip testing method, including: the CIS chip is enabled to output image data in an original format in a parallel or serial mode; analyzing the image data in the original format by adopting the digital image bad cluster statistical method to obtain an analysis result; and judging the quality of the CIS chip according to the analysis result and preset logic.
In order to solve the above technical problem, according to a third aspect of the present invention, there is provided an automatic integrated circuit tester, which adopts the above CIS chip testing method when testing CIS chips.
Compared with the prior art, the digital image bad cluster statistical method, the CIS chip test method and the integrated circuit automatic test machine provided by the invention comprise the following steps: decomposing the image data in the first format into a plurality of submatrices with preset sizes, and carrying out preset calculation on each submatrix and a nuclear matrix with preset sizes; wherein, the elements in the kernel matrix conform to the following rules: the result of multiplication of any two combinations comprising at least two elements is different. The configuration ensures that the number of the bad cluster type images with the preset size can be counted only by carrying out one-time scanning treatment and calculation on the original digital image, has the characteristics of rigorousness and high efficiency, and solves the problems of inaccurate counting result and low counting efficiency caused by the fact that the counting method of the bad clusters of the images in the prior art is not strict.
Drawings
Those of ordinary skill in the art will appreciate that the figures are provided for a better understanding of the present invention and do not constitute any limitation on the scope of the present invention. Wherein:
FIG. 1 is a flow chart of a method for digital image bad cluster statistics according to an embodiment of the invention;
fig. 2a is a schematic diagram of a 2 x 2 core matrix according to an embodiment of the present invention;
FIG. 2b is a schematic diagram of a core matrix of prime fill 5*5 in accordance with an embodiment of the invention;
FIG. 3 is a schematic diagram of a core matrix according to some embodiments of the invention;
FIG. 4 is a schematic diagram of converting image data in a primary format to image data in a secondary format according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a bad cluster distribution set according to one embodiment of the invention;
FIG. 6a is a schematic diagram of a first manner of decomposing image data in a first format into a plurality of sub-matrices 5*5 according to an embodiment of the invention;
FIG. 6b is a schematic diagram of a second way of decomposing image data in a first format into a plurality of sub-matrices 5*5 according to an embodiment of the invention;
FIG. 7 is a schematic diagram of a submatrix and a core matrix according to an embodiment of the present invention.
Detailed Description
The invention will be described in further detail with reference to the drawings and the specific embodiments thereof in order to make the objects, advantages and features of the invention more apparent. It should be noted that the drawings are in a very simplified form and are not drawn to scale, merely for convenience and clarity in aiding in the description of embodiments of the invention. Furthermore, the structures shown in the drawings are often part of actual structures. In particular, the drawings are shown with different emphasis instead being placed upon illustrating the various embodiments.
As used in this disclosure, the singular forms "a," "an," and "the" include plural referents, the term "or" are generally used in the sense of comprising "and/or" and the term "several" are generally used in the sense of comprising "at least one," the term "at least two" are generally used in the sense of comprising "two or more," and the term "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying any relative importance or number of features indicated. Thus, a feature defining "first," "second," "third," or "third" may explicitly or implicitly include one or at least two such features, the term "proximal" typically being one end proximal to the operator, the term "distal" typically being one end proximal to the patient, "one end" and "other" and "proximal" and "distal" typically referring to corresponding two portions, including not only the endpoints, the terms "mounted," "connected," "coupled," or "coupled" are to be construed broadly, e.g., as either a fixed connection, a removable connection, or as one piece; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communicated with the inside of two elements or the interaction relationship of the two elements. Furthermore, as used in this disclosure, an element disposed on another element generally only refers to a connection, coupling, cooperation or transmission between two elements, and the connection, coupling, cooperation or transmission between two elements may be direct or indirect through intermediate elements, and should not be construed as indicating or implying any spatial positional relationship between the two elements, i.e., an element may be in any orientation, such as inside, outside, above, below, or on one side, of the other element unless the context clearly indicates otherwise. The specific meaning of the above terms in the present invention can be understood by those of ordinary skill in the art according to the specific circumstances.
The invention provides a digital image bad cluster statistical method, a CIS chip test method and an integrated circuit automatic test machine, which are used for solving the problems of inaccurate statistical results and low statistical efficiency caused by the fact that the statistical method of the image bad clusters is not strict in the prior art.
The following description refers to the accompanying drawings.
Referring to fig. 1 to fig. 7, fig. 1 is a flow chart of a digital image bad cluster statistics method according to an embodiment of the invention; fig. 2a is a schematic diagram of a 2 x 2 core matrix according to an embodiment of the present invention; FIG. 2b is a schematic diagram of a core matrix of prime fill 5*5 in accordance with an embodiment of the invention; FIG. 3 is a schematic diagram of a core matrix according to some embodiments of the invention; FIG. 4 is a schematic diagram of converting image data in a primary format to image data in a secondary format according to an embodiment of the present invention; FIG. 5 is a schematic diagram of a bad cluster distribution set according to one embodiment of the invention; FIG. 6a is a schematic diagram of a first manner of decomposing image data in a first format into a plurality of sub-matrices 5*5 according to an embodiment of the invention; FIG. 6b is a schematic diagram of a second way of decomposing image data in a first format into a plurality of sub-matrices 5*5 according to an embodiment of the invention; FIG. 7 is a schematic diagram of a submatrix and a core matrix according to an embodiment of the present invention.
As shown in fig. 1, the present embodiment provides a digital image bad cluster statistical method, which includes:
s110, carrying out smoothing processing on the image data in the original format to obtain the image data in a smooth format, and subtracting the value of each pixel point in the image data in the original format from the value of the corresponding pixel point in the image data in the smooth format to obtain the image data in a second format;
it is to be understood that the image data may include PMD (pixel map definition). Specifically, the PMD may further include an original PMD, a rank PMD (rank refers to a smoothing algorithm), a flag PMD (flag refers to PMD data after identification), and so on. As an example, the original pmd may include image data in an original format; for another example, the rank pmd may include image data in a second format.
S120, setting a threshold value, performing assignment operation on the image data in the second format, assigning 1 to the point with the absolute value of the pixel point larger than or equal to the threshold value in the image data in the second format, and assigning 0 to the rest pixel points to obtain the image data in the first format; as an example, the flag pmd may include image data in a first format.
S130, obtaining a bad cluster distribution set, wherein the bad cluster distribution set comprises at least one bad cluster type image with a first format and a preset size; and carrying out preset calculation on each bad cluster type image and the nuclear matrix in the bad cluster distribution set to obtain a bad cluster type calculation result, and generating a lookup table between the corresponding bad cluster type image and the calculation result. (for details of the core matrix see step S140)
S140, decomposing the image data in the first format into a plurality of submatrices with preset sizes, and carrying out the preset calculation on each submatrix and the nuclear matrix with the preset size; wherein, the elements in the kernel matrix conform to the following rules: the results of multiplication of any two combinations comprising at least two elements are different;
s150, obtaining a calculation result of the submatrix and the nuclear matrix after the preset calculation, and comparing the calculation result with the lookup table to obtain bad cluster statistical information.
It should be understood that the elements in the kernel matrix may be understood as a complex number in the present specification, and for simplicity of calculation, a rational number or even an integer is often used; the preset size refers to the number of rows and columns of the matrix, for example, the preset size is 3×4, which means that the submatrix or the nuclear matrix is a matrix with 3 rows and 4 columns; further, the core matrix and the sub-matrix may be square matrices, i.e. the number of rows and columns is the same. The preset calculation should be understood broadly as a calculation that can be performed between two predefined matrices of the same size, and is set for the purpose of converting a relatively complex vector information (i.e. a matrix in the scope of the present specification) into a relatively easily handled scalar information (i.e. a number), and the specific embodiment may be a convolution calculation in the digital image processing field, that is, multiplication calculation of corresponding elements between two matrices, and then summation (or multiplication) of all products, or may be other calculation modes. When the value of the element in the sub-matrix is only 1 or 0, preferably, the preset calculation may include multiplying the element in the core matrix corresponding to the element with a value of 1 in the sub-matrix, where the corresponding concept may be understood as the element with the same number of rows and columns in both matrices.
In the embodiment, statistics can be completed by only carrying out one-time scanning treatment and calculation, and the method has the characteristics of rigorousness and high efficiency.
For "the elements in the kernel matrix follow the following law: any two features that are different from each other and that comprise a combination of at least two elements are described, for example, as follows, in one embodiment, a 2×2 kernel matrix is shown in fig. 2a, and the results of any combination and multiplication of the elements in the kernel matrix are exhaustive as follows: two elements are multiplied: 12 (2*6), 20 (2 x 10), 30 (2 x 15), 60 (6 x 10), 90 (6 x 15), 150 (10 x 15); three elements multiply: 120 (2 x 6 x 10), 180 (2 x 6 x 15), 300 (2 x 10 x 15), 900 (6 x 10 x 15); four elements are multiplied by 1800 (2 x 6 x 10 x 15). It is obvious that, in the above calculation results, any two calculation results are different, that is, the core matrix shown in fig. 2a conforms to the feature that "the result of multiplying any two combinations of elements in the core matrix, including at least two elements, is different". In some embodiments, other core matrix forms exist, for example, in a 2×2 core matrix, four elements are 0.2, 0.3, 0.5, and 0.7, respectively, which also can meet the characteristic that "any two elements in the core matrix contain a combination of at least two elements and the result of multiplication is different.
Although a core matrix satisfying the above characteristics can be obtained in a "try- (product to be multiplied) exhaustive-verification" manner for the core matrix of smaller size, the computational resources consumed for exhaustion and verification increase exponentially as the size of the core matrix increases, and the number of attempts to obtain a suitable core matrix also increases greatly, so that the above method is not a preferable solution in practical operation. On the other hand, when storing and calculating the decimal, irrational, and complex with imaginary numbers, the burden on the computer is also relatively large, so that it is possible to consider a case where the elements of the core matrix are all integers.
In a preferred embodiment, the kernel matrix is populated with varying prime numbers. Obviously, when the core matrix is filled with different prime numbers, the core matrix conforms to the characteristic that "the result of multiplication of any two combinations containing at least two elements in the core matrix is different". The specific proof process is common knowledge in the mathematical arts and is not described in the present specification.
Further, the kernel matrix is filled with a minimum of N prime numbers, where N is equal to the product of the number of rows of the preset size and the number of columns of the preset size. For example, a 5*5 core matrix is shown in fig. 2b (n= 5*5 =25 in this embodiment). The minimum N prime numbers are adopted for filling, so that the average expected value of the preset calculation is smaller, and the occupied calculation resources and storage resources in the operation process and the statistics process are smaller.
It should be understood that the above solution is a preferred solution, and those skilled in the art may also construct other core matrices according to the above concept, where the core matrix is consistent with the feature that "any two elements in the core matrix contain a combination of at least two elements, and the result of multiplication is different", for example, element 17 of the core matrix of 5*5 shown in fig. 2b is modified to 1.7; as another example, element 2 of the core matrix 5*5 shown in fig. 2b is modified to be 4, and obviously, such a scheme also accords with the characteristic that "any two elements in the core matrix contain the results of multiplying the combination of at least two elements are different", and the generation process is also far faster than the "try-and-exhaust-verification" mode, so that the generation process also has a better effect.
It should be understood that the size of the core matrix is not limited to 2 x 2 or 5*5, nor is it limited to a square matrix, and in other embodiments, other sizes of core matrices may be selected, and referring to fig. 3, in other embodiments, the core matrix may be one of the matrices shown in fig. 3.
In a preferred embodiment, the step S110 of obtaining the image data in the second format, please refer to fig. 4, includes: and carrying out smoothing processing on the image data in the original format to obtain the image data in the smoothed format, and subtracting the value of each pixel point in the image data in the original format from the value of the corresponding pixel point in the image data in the smoothed format to obtain the image data in the second format. The rule of the value of each element point in the original format image data can be set by a person skilled in the art through actual conditions and common knowledge, for example, the gray value of the shot image is assigned to each element of the original format image data. The smooth-format image data is obtained by performing smoothing processing on the original-format image data, and a specific smoothing algorithm can be used for carrying out average calculation on each pixel point by using surrounding pixel point assignment, or any one of the existing smoothing algorithms in the field can be selected according to actual needs. And subtracting the value of the corresponding pixel point in the image data in the smooth format from the value of each pixel point in the image data in the original format to obtain the image data in the second format. The image data in the second format can be understood as a large sparse matrix, in which non-zero elements are potential outliers.
Then, the image data in the first format is obtained in step S120, where the pixel points of the image data are assigned with 0 or 1, where the normal pixel points are assigned with 0, and the abnormal pixel points are assigned with 1. Step S120 includes: setting a threshold value, performing assignment operation on image data in a second format, assigning a point with an absolute value greater than or equal to the threshold value in the image data in the second format to be 1, and assigning the rest pixel points to be 0 to obtain the image data in the first format. For example, in an embodiment, the values of the non-zero elements in the image data in the second format are 1, 5, -6, 7, and-8, by setting the threshold 6, the elements with values of 1 and 5 are assigned to 0, the elements with values of-6, 7, and-8 are assigned to 1, and the remaining elements that were originally 0 can be assigned to 0 again, which can simplify the flow and avoid operation. The specific threshold selection process can be determined according to actual needs, for example, a smaller threshold is set under the working condition that the abnormal points of the image data need to be counted more strictly, and a larger threshold is set under the working condition that the abnormal points of the image data need to be counted more loosely. Of course, the threshold value is set in consideration of eliminating the influence of noise.
Then, step S130 is performed: obtaining a bad cluster distribution set, wherein the bad cluster distribution set comprises at least one bad cluster type image with the preset size in the first format; and carrying out the preset calculation on each bad cluster type image and the core matrix in the bad cluster distribution set to obtain a bad cluster type calculation result, and generating a lookup table between the corresponding bad cluster type image and the calculation result.
In this embodiment, the sub-matrix into which bad cluster statistics is counted includes at least two bad points (abnormal points), and the matrix of a single bad point does not count bad cluster types; wherein bad points (outliers) refer to non-zero elements in the sub-matrix of the first format. The bad cluster distribution set may be set according to experience and actual needs, for example, when the preset size is n×n, it may be considered that only the number of outliers is 2 to (n×n) are added to the bad cluster distribution set, and further, when the preset size is 5*5, it may be considered that only the number of outliers is 2 to 25 are added to the bad cluster distribution set. For another example, only the case where the number of abnormal points is 3 to 4 is added to the bad cluster distribution set, and the rest cases are not added to the bad cluster distribution set. In the subsequent processing, the processing may be performed according to actual needs, for example, when the preset size is 5*5, the statistical objective is to count the number of abnormal points to be 3-4, and scheme 1 may be selected: the set bad cluster distribution set comprises the situation that the number of the abnormal points is 2-25, but in the subsequent statistics, only the situation that the number of the abnormal points is 3-4 is counted, and other situations are eliminated; the scheme 2 can also be selected, the set bad cluster distribution set only comprises the situation that the number of the abnormal points is 3-4, and all situations contained in the lookup table are counted in the follow-up statistics; in the scheme 3, the bad cluster distribution set may be configured to include only 3 to 4 abnormal points, or may be configured to include 2 to 25 abnormal points, but in the subsequent statistics, the preset calculation is performed only when the number of abnormal points is 3 to 4, and other submatrices, such as 2 or 5 points, are not calculated. The foregoing is merely an exemplary expression, and in practice, the bad cluster distribution set to be defined may be arbitrarily selected according to actual needs. Referring to fig. 5, fig. 5 is a schematic diagram of a bad cluster distribution set according to an embodiment of the invention. In other embodiments, the bad cluster distribution set may not be the same as that of FIG. 5. So configured, the subsequent computational burden can be reduced. It should be understood that a lookup table may not be generated, and the bad cluster type image corresponding to each calculation result may be back-pushed in real time in the statistical process. But the scheme of generating the lookup table is favorable for saving calculation resources in the statistical process, and is a preferable scheme.
In step S140, the specific splitting manners of the sub-matrices include at least two kinds, and the first splitting manner is to perform non-repeated splitting, please refer to fig. 6a, in which in fig. 6a, a data image in the first format is decomposed into 4 sub-matrices 5*5, and each sub-matrix does not include repeated elements therebetween. The second splitting method is to split row by column, that is, each adjacent submatrix is shifted by only one row or one column, please refer to fig. 6b, in which in fig. 6b, a data image in the first format is decomposed into 36 submatrices of 5*5. Of course, other ways of splitting are possible, as well, due to special requirements under special conditions. It should be understood that, in the first splitting manner, if the number of rows or columns of the data image in the first format cannot be exactly divided by the number of rows or columns in the preset size, the following scheme may be adopted to solve the problem: the incomplete submatrices which do not meet the preset size are firstly split, then 0 elements are used for complement, and the complement positions of the 0 elements can be one or more of the upper side, the lower side, the left side and the right side of the incomplete matrix. Other ways of supplementing the data image may be adopted, for example, supplementing adjacent elements in the original data image with a incomplete matrix, and those skilled in the art can set the data image according to actual needs. Since the data image in the first format is set to include only the element of 0 or 1, in the foregoing discussion, it has been mentioned that when the value of the element in the submatrix is only 1 or 0, a preferred preset calculation mode may be selected, and in this embodiment, the preset calculation is that the elements in the kernel matrix corresponding to the element with the value of 1 in the submatrix are multiplied to obtain a product.
Referring to fig. 7, for example, in one embodiment, one of the submatrices is shown on the left side of fig. 7 and the core matrix is shown on the right side of fig. 7. The elements in the core matrix corresponding to the elements assigned 1 in the submatrix are respectively 2, 13, 23, 29, 31, 41, 47, 53, 73, 79, 89, and the result of the preset calculation of the submatrix and the core matrix is 28181130952459700 (i.e. 2×13×23×29×31×41×47×53×73×79×89).
In step S150, the digital image bad cluster statistical method includes: and obtaining a calculation result after the sub-matrix and the nuclear matrix are subjected to preset calculation, and comparing the calculation result with the lookup table to obtain bad cluster statistical information. Specific statistical information includes, for example, counting the number of occurrences of each different bad cluster, or first classifying the bad cluster, and then counting the number of occurrences of each classification, which may be data of interest to other operators.
The embodiment also provides a CIS chip testing method, which comprises the following steps:
the CIS chip is enabled to output image data in an original format in a parallel or serial mode;
analyzing the image data in the original format by adopting the digital image bad cluster statistical method to obtain an analysis result;
and judging the quality of the CIS chip according to the analysis result and preset logic.
The method for the CIS chip to output the image data in the original format may be configured by those skilled in the art according to actual circumstances, for example, the CIS chip is irradiated with a test light source, etc. The preset logic can be configured by a person skilled in the art according to actual conditions, for example, for products with higher quality requirements, the preset logic is set more strictly, and for products with lower quality requirements, the preset logic is set more loosely; for another example, for a CIS chip with a specific requirement, the preset logic is set to determine that the chip is failed when the number of occurrences of the specific abnormal point distribution pattern is greater than 1.
The embodiment also provides an automatic integrated circuit testing machine, which adopts the CIS chip testing method when testing the CIS chip. The integrated circuit automatic tester is used for testing the rack structure of the CIS chip, the communication interface and calling the internal logic of the CIS chip to work, and can be set according to common knowledge by a person skilled in the art, and other components of the integrated circuit automatic tester can also be set according to common knowledge in the art, and are not described in detail herein.
The CIS chip testing method and the integrated circuit automatic testing machine both adopt the digital image bad cluster statistical method, so the CIS chip testing method and the integrated circuit automatic testing machine have the same rigorous and efficient beneficial effects.
In summary, in the digital image bad cluster statistical method, the CIS chip test method and the integrated circuit automatic test machine provided in the embodiment, the digital image bad cluster statistical method can complete statistics only by performing scanning processing and calculation once, and has the characteristics of rigorousness and high efficiency.
The foregoing description is only illustrative of the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention in any way, and any changes and modifications made by those skilled in the art in light of the foregoing disclosure will be deemed to fall within the scope and spirit of the present invention.
Claims (7)
1. The digital image bad cluster statistical method is characterized by comprising the following steps of: decomposing the image data in the first format into a plurality of submatrices with preset sizes, and carrying out preset calculation on each submatrix and a nuclear matrix with preset sizes; wherein, the elements in the kernel matrix conform to the following rules: the results of multiplication of any two combinations comprising at least two elements are different;
the first format is that 0 or 1 is used for assigning the pixel points of the image data, wherein 0 is used for assigning the normal pixel points, and 1 is used for assigning the abnormal pixel points;
meanwhile, the fact that the elements in the submatrices with equal row numbers and equal column numbers and the elements in the kernel matrix have corresponding relations is met, and the preset calculation comprises the steps of multiplying the elements in the kernel matrix corresponding to the elements with the value of 1 in the submatrices;
the digital image bad cluster statistical method comprises the following steps: after obtaining the image data in the first format, obtaining a bad cluster distribution set, wherein the bad cluster distribution set comprises at least one bad cluster type image with the preset size in the first format; carrying out the preset calculation on each bad cluster type image and the nuclear matrix in the bad cluster distribution set to obtain a bad cluster type calculation result, and generating a lookup table between the corresponding bad cluster type image and the calculation result;
the digital image bad cluster statistical method comprises the following steps: and obtaining a calculation result of the submatrix and the nuclear matrix after the preset calculation, and comparing the calculation result with the lookup table to obtain bad cluster statistical information.
2. The digital image bad cluster statistical method of claim 1, wherein the kernel matrices are populated with different prime numbers.
3. The digital image bad cluster statistical method of claim 1, wherein the kernel matrix is filled with a minimum of N prime numbers, where N is equal to a product of the number of rows of the preset size and the number of columns of the preset size.
4. The digital image bad cluster statistical method of claim 1, wherein the digital image bad cluster statistical method comprises: setting a threshold value, performing assignment operation on image data in a second format, assigning a point with an absolute value greater than or equal to the threshold value in the image data in the second format to be 1, and assigning the rest pixel points to be 0 to obtain the image data in the first format.
5. The digital image bad cluster statistical method of claim 4, wherein the digital image bad cluster statistical method comprises: and carrying out smoothing processing on the image data in the original format to obtain the image data in the smoothed format, and subtracting the value of each pixel point in the image data in the original format from the value of the corresponding pixel point in the image data in the smoothed format to obtain the image data in the second format.
6. A CIS chip testing method, comprising:
the CIS chip is enabled to output image data in an original format in a parallel or serial mode;
analyzing the image data in the original format by adopting the digital image bad cluster statistical method according to any one of claims 1 to 5 to obtain an analysis result;
and judging the quality of the CIS chip according to the analysis result and preset logic.
7. An automatic integrated circuit tester, wherein the automatic integrated circuit tester adopts the CIS chip testing method according to claim 6 when testing CIS chips.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011479362.5A CN112596965B (en) | 2020-12-14 | 2020-12-14 | Digital image bad cluster statistical method and integrated circuit automatic tester |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011479362.5A CN112596965B (en) | 2020-12-14 | 2020-12-14 | Digital image bad cluster statistical method and integrated circuit automatic tester |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112596965A CN112596965A (en) | 2021-04-02 |
| CN112596965B true CN112596965B (en) | 2024-04-09 |
Family
ID=75195948
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011479362.5A Active CN112596965B (en) | 2020-12-14 | 2020-12-14 | Digital image bad cluster statistical method and integrated circuit automatic tester |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112596965B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115713474B (en) * | 2022-12-30 | 2024-03-08 | 安徽光智科技有限公司 | Infrared image correction method and system and electronic equipment |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995015530A1 (en) * | 1993-11-30 | 1995-06-08 | Polaroid Corporation | Image coding by use of discrete cosine transforms |
| CN105306843A (en) * | 2015-10-20 | 2016-02-03 | 凌云光技术集团有限责任公司 | Dead pixel processing method and system for image sensor |
| CN105654126A (en) * | 2015-12-29 | 2016-06-08 | 华为技术有限公司 | Computing equipment, kernel matrix evaluation method and multi-kernel learning method |
| CN108200439A (en) * | 2013-06-14 | 2018-06-22 | 浙江大学 | The method and digital signal converting method and device of raising digital signal conversion performance |
| CN109255434A (en) * | 2018-10-15 | 2019-01-22 | 旺微科技（上海）有限公司 | The dispatching method and device of computing resource in a kind of convolutional neural networks |
| CN109671081A (en) * | 2018-12-25 | 2019-04-23 | 凌云光技术集团有限责任公司 | A kind of bad cluster statistical method and device based on FPGA look-up table |
| CN109784372A (en) * | 2018-12-17 | 2019-05-21 | 北京理工大学 | A kind of objective classification method based on convolutional neural networks |
| CN110020678A (en) * | 2019-03-25 | 2019-07-16 | 联想(北京)有限公司 | A kind of data processing method, electronic equipment and computer storage medium |
| CN110188812A (en) * | 2019-05-24 | 2019-08-30 | 长沙理工大学 | A kind of multicore clustering method of quick processing missing isomeric data |
| CN111767956A (en) * | 2020-06-30 | 2020-10-13 | 苏州科达科技股份有限公司 | Image tampering detection method, electronic device, and storage medium |
| CN111932437A (en) * | 2020-10-10 | 2020-11-13 | 深圳云天励飞技术股份有限公司 | Image processing method, image processing device, electronic equipment and computer readable storage medium |
-
2020
- 2020-12-14 CN CN202011479362.5A patent/CN112596965B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995015530A1 (en) * | 1993-11-30 | 1995-06-08 | Polaroid Corporation | Image coding by use of discrete cosine transforms |
| CN108200439A (en) * | 2013-06-14 | 2018-06-22 | 浙江大学 | The method and digital signal converting method and device of raising digital signal conversion performance |
| CN105306843A (en) * | 2015-10-20 | 2016-02-03 | 凌云光技术集团有限责任公司 | Dead pixel processing method and system for image sensor |
| CN105654126A (en) * | 2015-12-29 | 2016-06-08 | 华为技术有限公司 | Computing equipment, kernel matrix evaluation method and multi-kernel learning method |
| CN109255434A (en) * | 2018-10-15 | 2019-01-22 | 旺微科技（上海）有限公司 | The dispatching method and device of computing resource in a kind of convolutional neural networks |
| CN109784372A (en) * | 2018-12-17 | 2019-05-21 | 北京理工大学 | A kind of objective classification method based on convolutional neural networks |
| CN109671081A (en) * | 2018-12-25 | 2019-04-23 | 凌云光技术集团有限责任公司 | A kind of bad cluster statistical method and device based on FPGA look-up table |
| CN110020678A (en) * | 2019-03-25 | 2019-07-16 | 联想(北京)有限公司 | A kind of data processing method, electronic equipment and computer storage medium |
| CN110188812A (en) * | 2019-05-24 | 2019-08-30 | 长沙理工大学 | A kind of multicore clustering method of quick processing missing isomeric data |
| CN111767956A (en) * | 2020-06-30 | 2020-10-13 | 苏州科达科技股份有限公司 | Image tampering detection method, electronic device, and storage medium |
| CN111932437A (en) * | 2020-10-10 | 2020-11-13 | 深圳云天励飞技术股份有限公司 | Image processing method, image processing device, electronic equipment and computer readable storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 二维数字图像分形维数的计算方法;彭瑞东, 谢和平, 鞠杨;中国矿业大学学报(第01期);全文 * |
| 彭瑞东,谢和平,鞠杨.二维数字图像分形维数的计算方法.中国矿业大学学报.2004,(第01期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112596965A (en) | 2021-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111401512B (en) | Method and system for convolution in neural networks with variable expansion rate | |
| CN112596965B (en) | Digital image bad cluster statistical method and integrated circuit automatic tester | |
| US20210224954A1 (en) | Image processing method and device, storage medium and computer device | |
| US20240211535A1 (en) | Sparse matrix accelerated computing method and apparatus, device, and medium | |
| EP4071615A1 (en) | Quantum error correction decoding system and method, fault-tolerant quantum error correction system, and chip | |
| CN109255390B (en) | Training image preprocessing method and module, discriminator and readable storage medium | |
| CN116362630B (en) | Tin paste printer management method, system and medium based on Internet of things | |
| CN110378313A (en) | Cell mass recognition methods, device and electronic equipment | |
| CN109543772B (en) | Data set automatic matching method, device, equipment and computer readable storage medium | |
| CN109325061A (en) | A kind of method for processing abnormal data and device based on probability distribution | |
| CN106778550B (en) | Face detection method and device | |
| CN112818632B (en) | Analysis method and device for pattern density of chip and electronic equipment | |
| CN117726322A (en) | Intelligent management method and system for probe test equipment | |
| CN113094415B (en) | Data extraction method, data extraction device, computer readable medium and electronic equipment | |
| CN112712181A (en) | Model construction optimization method, device, equipment and readable storage medium | |
| CN112084728B (en) | Pix2 pix-based PCB gray image coloring method and system | |
| CN115174421A (en) | Network fault prediction method and device based on self-supervision unwrapping hypergraph attention | |
| WO2021122336A1 (en) | Apparatus and method for generating higher-level features | |
| US10755022B2 (en) | Electronic apparatus and layout method for integrated circuit | |
| CN113495786B (en) | Image convolution processing method and electronic equipment | |
| EP4432136A1 (en) | Computer-implemented data processing method, and measurement system | |
| CN113884857B (en) | Chip, chip pressure testing method and device, electronic equipment and storage medium | |
| US20240202875A1 (en) | Method of detecting wafer defects based on singularly valuable decomposition | |
| US11852673B1 (en) | Method for generating chip probing wafer map | |
| WO2022014392A1 (en) | Data processing device, data processing system, data processing method, and data processing program |
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 |