CN103927312A - Automatic classification method and system for failure information of CIS (contact image sensor) - Google Patents
Automatic classification method and system for failure information of CIS (contact image sensor) Download PDFInfo
- Publication number
- CN103927312A CN103927312A CN201310015074.8A CN201310015074A CN103927312A CN 103927312 A CN103927312 A CN 103927312A CN 201310015074 A CN201310015074 A CN 201310015074A CN 103927312 A CN103927312 A CN 103927312A
- Authority
- CN
- China
- Prior art keywords
- module
- pattern
- positional information
- file
- fail message
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Abstract
The invention provides an automatic classification method and system for failure information of a CIS (contact image sensor). The automatic classification method includes the steps: outputting a detection result picture file which is a BMP (bitmap) picture file by the CIS; reading the detection result picture file and extracting the failure information. The specific process includes: switching the BMP picture file into a hexadecimal file; finding the gray value range of failure bits by a hierarchical cluster analysis method according to the characteristic of pixel color jump of normal bits and the failure bits; switching the failure information into a result file in a standard text format; automatically classifying the failure information in the result file according to a user-defined failure mode; displaying an automatic classification result. The problem of difficulty in automatic classification analysis for CIS product image testing results is rapidly solved, and the problems of low efficiency and inaccuracy of statistics and classification manually performed with eyes for a long time are avoided.
Description
Technical field
The invention belongs to semiconductor detection technique field, relate to a kind of fail message automatic classification method and system of CIS imageing sensor.
Background technology
Cmos image sensor is along with the development of mobile phone camera is with rapid changepl. never-ending changes and improvements.From 20K pixel in one's early years 32M pixel till now, different company, university and research institution are all ceaselessly developing cmos image sensor all over the world, have new technology to produce every year.Along with constantly reducing of Pixel Design, it has all produced new challenge to the production technology of cmos image sensor and quality.In order to guarantee the quality of product, each image sensor needs test verification, and draws corresponding BMP picture, can find out the graphic feature of various failure modes on problematic product from BMP picture.
In order to obtain the information of cmos image sensor failure mechanism, need to find the position of each inefficacy pixel on problematic cmos image sensor chip.The method of determining at present inefficacy location of pixels is manually to utilize naked eyes to add up the fail bit obtaining, and makes chart, classification analysis then, thus find the mechanism of its inefficacy.As shown in Figure 1, it is the test result schematic diagram of a CIS product, wherein has the BMP fail message of single-point, and the BMP fail message of strip.The type of the fail message shown in Fig. 1 is fairly simple, easy artificial judgment, but when the type of fail message and quantity are all very many, utilize naked eyes to judge that to fail message classification is obviously unpractical, the greatest problem that is said method is: need to spend a large amount of mental and time of slip-stick artist, and often occur the situations such as drop-out, erroneous judgement, so said method is along with the capacity of CIS sensor increases and becomes difficult to achieve.
Summary of the invention
The shortcoming of prior art in view of the above, the object of the present invention is to provide a kind of fail message automatic classification method and system of CIS imageing sensor, poor efficiency and coarse problem of for solving prior art, manually the fail message of CIS imageing sensor being added up and being sorted out.
For achieving the above object and other relevant objects, the invention provides a kind of fail message automatic classification method and system of CIS imageing sensor, the fail message automatic classification method of wherein said CIS imageing sensor comprises: CIS imageing sensor output detections result picture file; Described testing result picture file is BMP picture file; Understand described testing result picture file, extract fail message, detailed process comprises: convert described BMP picture file to hex file; The feature that has saltus step according to the pixel color of normal position and fail bit, finds the gray-scale value scope of fail bit by hierarchical cluster analysis method; Described fail message is converted to the destination file of received text form; According to self-defined failure mode, the fail message in described destination file is carried out to robotization classification; Show robotization classification result.
Preferably, before the described gray-scale value scope step that finds fail bit by hierarchical cluster analysis method, first described hex file is carried out to frame filtration treatment, the detailed process of filtration treatment is: reject full line in described hex file or permutation black picture element and surpass 90% frame.
Preferably, the specific implementation process that described fail message is converted to the destination file of received text form comprises: read the positional information of the gray-scale value scope that meets described fail bit, described positional information is output into the destination file of received text form.
Preferably, described self-defined failure mode comprises: with shape definition failure mode, comprise single-point pattern, two point pattern, line pattern, alignment pattern, intersecting lens pattern, region mode.
Preferably, the mode that described robotization is sorted out comprises: according to principle from big to small, by the order of face, line, point, classify successively, specific implementation process is: judge whether the positional information in described destination file belongs to region mode, if be referred to region mode; Otherwise continue to judge whether described positional information belongs to line pattern, alignment pattern or intersecting lens pattern, if be referred to corresponding line pattern, alignment pattern or intersecting lens pattern; Otherwise continue to judge whether described positional information belongs to single-point pattern or two point pattern, if be referred to corresponding single-point pattern or two point pattern.
Preferably, the mode that described robotization is sorted out comprises: the feature that is presented as face, row, column, point according to described fail message, positional information in described destination file is pressed to line direction sequence, sequence in column direction again, then the positional information after sequence is carried out to cluster analysis, the failure mode that finds each positional information to belong to.
Preferably, the implementation procedure that result is sorted out in described demonstration robotization comprises: adopt failure mode interface or/and statistics figure surface and interface or/and sensor failure arrangement information interface or/and failure mode and bitmap display interface show that described robotization sorts out result.
The fail message automatic classification system of described CIS imageing sensor comprises: file layout modular converter, fail message extraction module, received text formatted output module, automatic clustering module; The testing result picture file that described file layout modular converter is exported in order to receive CIS imageing sensor, and convert described testing result picture file to hex file; Described fail message extraction module is connected with described file layout modular converter, in order to there to be the feature of saltus step according to the pixel color of normal position and fail bit, finds the gray-scale value scope of fail bit by hierarchical cluster analysis method; Described received text formatted output module is connected with described fail message extraction module, in order to described fail message is converted to the destination file of received text form; Described automatic clustering module, is connected with described received text formatted output module, in order to the fail message in described destination file is carried out to robotization classification according to self-defining failure mode.
Preferably, the fail message automatic classification system of described CIS imageing sensor also comprises: filtration module, display module; Described filtration module is connected respectively with described fail message extraction module with described file layout modular converter, in order to full line in described hex file or permutation black picture element are surpassed to 90% frame filtering; Described display module is connected with described automatic clustering module, in order to show robotization classification result.
Preferably, described self-defined failure mode comprises: with shape definition failure mode, comprise single-point pattern, two point pattern, line pattern, alignment pattern, intersecting lens pattern, region mode.
Preferably, described automatic clustering module comprises: the first judge module, the firstth, and execution module, the first no execution module, the second judge module, the secondth, execution module, the second no execution module, the 3rd judge module, the 3rd are execution modules; Described the first judge module is connected with described received text formatted output module, in order to judge whether the positional information in described destination file belongs to region mode; Described first is that execution module is connected with described the first judge module, while belonging to region mode in order to the positional information in described destination file, described positional information is referred to region mode; Described the first no execution module is connected with described the first judge module, carries out one second judge module while not belonging to region mode in order to the positional information in described destination file; Described the second judge module, is connected with described the first no execution module, in order to continue judging whether described positional information belongs to line pattern, alignment pattern or intersecting lens pattern; Described second is that execution module is connected with described the second judge module, described positional information is referred to corresponding line pattern, alignment pattern or intersecting lens pattern when belonging to line pattern, alignment pattern or intersecting lens pattern in described positional information; Described the second no execution module is connected with described the second judge module, carries out one the 3rd judge module when not belonging to line pattern, alignment pattern or intersecting lens pattern in described positional information; Described the 3rd judge module is connected with described the second no execution module, in order to continue judging whether described positional information belongs to single-point pattern or two point pattern; The described the 3rd is that execution module is connected with described the 3rd judge module, described positional information is referred to corresponding single-point pattern or two point pattern when belonging to single-point pattern or two point pattern in described positional information.
Preferably, described automatic clustering module comprises: order module, cluster analysis module; Described order module is connected with described received text formatted output module, in order to be presented as the feature of face, row, column, point according to described fail message, positional information in described destination file is pressed to line direction sequence, then sequence in column direction, the positional information after sequence obtained; Described cluster analysis module is connected with described order module, in order to the positional information after described sequence is carried out to cluster analysis, and the failure mode that finds each positional information to belong to.
As mentioned above, fail message automatic classification method and the system of CIS imageing sensor of the present invention, there is following beneficial effect: the present invention has solved the difficult problem that CIS product image test result is carried out to automatic clustering analysis fast, avoided poor efficiency and the out of true problem of manually with eyes, adding up and sorting out for a long time; In addition, the present invention has realized the BMP picture file of CIS imageing sensor output has been understood accurately, has realized the automatic partition alanysis of image feature from the angle of system, has greatly improved production efficiency.
Accompanying drawing explanation
Fig. 1 is the test result schematic diagram of an existing CIS product.
Fig. 2 is the schematic flow sheet of the fail message automatic classification method of CIS imageing sensor of the present invention.
Fig. 3 is the further detailed process schematic diagram of the fail message automatic classification method of CIS imageing sensor of the present invention.
Fig. 4 is a kind of realization flow schematic diagram that in the fail message automatic classification method of CIS imageing sensor of the present invention, step is sorted out in robotization.
Fig. 5 is the another kind of realization flow schematic diagram that in the fail message automatic classification method of CIS imageing sensor of the present invention, step is sorted out in robotization.
Fig. 6 is the structural representation of the fail message automatic classification system of CIS imageing sensor of the present invention.
Fig. 7 is a kind of structural representation of automatic clustering module in the fail message automatic classification system of CIS imageing sensor of the present invention.
Fig. 8 is the another kind of structural representation of automatic clustering module in the fail message automatic classification system of CIS imageing sensor of the present invention.
Element numbers explanation
1 file layout modular converter
2 filtration modules
3 fail message extraction modules
4 received text formatted output modules
5 automatic clustering modules
6 display modules
51 first judge modules
52 first is execution module
53 first no execution modules
54 second judge modules
55 second is execution module
56 second no execution modules
57 the 3rd judge modules
58 the 3rd is execution module
51A order module
52A cluster analysis module
Embodiment
Below, by specific instantiation explanation embodiments of the present invention, those skilled in the art can understand other advantages of the present invention and effect easily by the disclosed content of this instructions.The present invention can also be implemented or be applied by other different embodiment, and the every details in this instructions also can be based on different viewpoints and application, carries out various modifications or change not deviating under spirit of the present invention.
Refer to accompanying drawing.It should be noted that, the diagram providing in the present embodiment only illustrates basic conception of the present invention in a schematic way, satisfy and only show with assembly relevant in the present invention in graphic but not component count, shape and size drafting while implementing according to reality, during its actual enforcement, kenel, quantity and the ratio of each assembly can be a kind of random change, and its assembly layout kenel also may be more complicated.
Below in conjunction with embodiment and accompanying drawing, the present invention is described in detail.
Embodiment
The present embodiment provides a kind of fail message automatic classification method of CIS imageing sensor, and as shown in Figure 2, this automatic classification method comprises:
CIS imageing sensor output detections result picture file.Further, the abbreviation that described testing result picture file is BMP(bitMap) picture file can be also the picture file of other types.BMP is the integrated of a certain width and pixel highly, has 2 color bitmaps, 16 color bitmaps, 256 color bitmaps and 24 bitmaps.What semiconductor testing apparatus was generally exported is the black and white picture of 256 color bitmaps, detect by an unaided eye and can find, every have the place of fail bit to show outside the color that different gray scales are more black, and other place shows the whiter color of different gray scales, shown in Figure 1.
Understand described testing result picture file, extract fail message, detailed process comprises: convert described check result picture file to hex file, have the feature of saltus step according to the pixel color of normal position and fail bit, by hierarchical cluster analysis method, extract fail message.
The present embodiment be take BMP picture file as example, and the implementation procedure of this step is described in detail.Described BMP picture file converts to after hex file, and in described hex file, the gray-scale value scope of pixel is that 00(is black)～FF(is white), totally 256 gray shade scales; The feature that has saltus step according to the pixel color of normal position and fail bit, by hierarchical cluster analysis method, can find the gray-scale value scope of fail bit is 00～2F, and and the gray-scale value scope of normal be 50～FF.The color of different pictures and gray scale are different, on picture, although fail bit color is more black, but the color value of its pixel or different, but the color of normal position and fail bit is but to have very large saltus step, the method by hierarchical cluster analysis just can find the greyscale color of fail bit very soon.Through a large amount of data analyses, find, the gray-scale value scope of general fail bit is 00～2F, and the gray-scale value scope of other normal position is 50～FF.
In order to improve the efficiency of the method for the invention, also 00～FF256 data value can be divided into 32 grades, also with regard to saying, 8 continuous hexadecimal values are considered as to a value, so just have 1, 2, 32 grade points, color (the being gray scale) scope of the color based on fail bit (being gray scale) and other normal position has the feature of a large saltus step, so will obtain two large clusters with the algorithm of cluster analysis, that is to say data have been divided into two groups, first group of data is at grade 1-6(00～2F) in, second group of data is at grade 11-32(50～FF), intermediate grade 7-10(30～4F) be a large tomography.
Further, as shown in Figure 3, before the described gray-scale value scope step that finds fail bit by hierarchical cluster analysis method, first described hex file is carried out to frame filtration treatment, the detailed process of filtration treatment is: reject full line in described hex file or permutation black picture element and surpass 90% frame.Therefore due to the factors such as background that picture is taken pictures, the periphery of some pictures looks the same with the color of fail bit, all presents black, these must be presented to black but is not that the perimeter filter of fail bit is gone out.The black picture element of above-mentioned periphery all embodies the feature of continuous full line or permutation, and in order to filter fast, the present invention need adopt simple and effective algorithm to be rejected.Described in the present embodiment the full line of periphery or permutation black picture element to be surpassed continuously to 90% frame filters out be exactly the method for demand according to the invention.
Described fail message is converted to the destination file of received text form, specific implementation process comprises: read the positional information of the gray-scale value scope that meets described fail bit, described positional information is output into the destination file of received text form.In order to read accurately the positional information of fail bit, must grasp the structure of BMP picture file, the most important thing is that a pixel of picture is by several value representations, the height of image has how many pixels, and wide have how many pixels.The file (being the destination file of described received text form) of output must be the normative document that common system needs.
According to self-defined failure mode, the fail message in described destination file is carried out to robotization classification.Wherein, described self-defined failure mode comprises with shape definition failure mode or with other formal definition failure modes; the define styles of this failure mode can carry out according to actual conditions, thus protection scope of the present invention be not limited to that the present embodiment exemplifies in this kind of mode of shape definition.Some basic failure modes, by shape definition, as the fail bit of single-point, are called Single Bit; Two continuous fail bits (Fail Bit), are called Twin Bit; Fail bit forms a continuous line of line direction, referred to as WL; Fail bit forms a continuous line of column direction, referred to as BL; Fail bit forms two intersecting lenses, is called Cross; Or fail bit forms a region, is called Block etc.The described failure mode with shape definition comprises: single-point pattern, two point pattern, line pattern, alignment pattern, intersecting lens pattern, region mode etc.Robotization sort out process be first by user, the feature of failure mode to be defined, then system goes to check according to user-defined failure mode, finds the fail message being consistent with user-defined failure mode (Fail Pattern).
Further, as shown in Figure 4, a kind of implementation that described robotization is sorted out comprises: according to principle from big to small, by the order of face, line, point, classify successively, specific implementation process is: judge whether the positional information in described destination file belongs to region mode, if be referred to region mode; Otherwise continue to judge whether described positional information belongs to line pattern, alignment pattern or intersecting lens pattern, if be referred to corresponding line pattern, alignment pattern or intersecting lens pattern; Otherwise continue to judge whether described positional information belongs to single-point pattern or two point pattern, if be referred to corresponding single-point pattern or two point pattern.This step filters out the data that detected, then according to detection principle from big to small, from face → line → the method for inspection carry out classification analysis, solved a difficult problem for classification searching efficiency.
Further again, as shown in Figure 5, the practical implementation of another kind that described robotization is sorted out comprises: the feature that is presented as face, row, column, point according to described fail message, positional information in described destination file is pressed to line direction sequence, sequence in column direction again, then the positional information after sequence is carried out to cluster analysis, the failure mode that finds each positional information to belong to.Although the data of failure mode are huge, but the definition of failure mode is presented as face, row, column, point,, there is no the definition of irregular failure mode, so the implementation that the robotization described in this step is sorted out is to realize according to the definition feature of failure mode.
Show robotization classification result.The implementation that result is sorted out in described demonstration robotization has multiple, can select according to actual needs, as this display mode comprises: adopt failure mode interface or/and statistics figure surface and interface or/and sensor failure arrangement information interface or/and failure mode and bitmap display interface show as described in robotization sort out result.
The present invention is by the image information of accurate reading CIS imageing sensor output, solution is read the fail message in this image information, and fail message is converted to the destination file of conventional text formatting, then the fail message in destination file is carried out the classification analysis of robotization according to self-defining failure mode, finally with a good instrument, show classification results, solve fast the difficult problem that CIS product image test result is carried out to automatic clustering analysis, avoided poor efficiency and the out of true problem of manually with eyes, adding up and sorting out for a long time.The present invention has realized the BMP picture file of CIS imageing sensor output has been understood accurately, has realized the automatic partition alanysis of image feature from the angle of system, has greatly improved production efficiency.
The present embodiment also provides a kind of fail message automatic classification system of CIS imageing sensor, this automatic classification system can be realized automatic classification method of the present invention, as shown in Figure 6, this automatic classification system comprises: file layout modular converter 1, filtration module 2, fail message extraction module 3, received text formatted output module 4, automatic clustering module 5, display module 6.Described filtration module 2 is connected respectively with described fail message extraction module 3 with described file layout modular converter 1, described received text formatted output module 4 is connected with described fail message extraction module 3, described automatic clustering module 5 is connected with described received text formatted output module 4, and described display module 6 is connected with described automatic clustering module 5.
The testing result picture file that described file layout modular converter 1 is exported in order to receive CIS imageing sensor, and convert described testing result picture file to hex file.Described testing result picture file can adopt BMP picture file.Described BMP picture file is converted to after hex file, and in described hex file, the gray-scale value scope of pixel is 00～FF.
Described filtration module 2 is in order to surpass 90% frame filtering by full line in described hex file or permutation black picture element.
Described fail message extraction module 3, in order to there to be the feature of saltus step according to the pixel color of normal position and fail bit, finds the gray-scale value scope of fail bit by hierarchical cluster analysis method.Described fail message extraction module 3 has the feature of saltus step according to the pixel color of normal position and fail bit, by hierarchical cluster analysis method, finding the gray-scale value scope of fail bit is 00～2F, and the gray-scale value scope of normal position is 50～FF.
Described received text formatted output module 4 is in order to convert described fail message to the destination file of received text form.Particularly, described received text formatted output module 4 reads the positional information of the gray-scale value scope that meets described fail bit, described positional information is output into the destination file of received text form.
Described automatic clustering module 5 is in order to carry out robotization classification according to self-defining failure mode to the fail message in described destination file.Wherein, described self-defined failure mode includes but not limited to shape definition failure mode.Failure mode with shape definition comprises single-point pattern, two point pattern, line pattern, alignment pattern, intersecting lens pattern, region mode.
Further, a kind of implementation structure of described automatic clustering module 5 as shown in Figure 7, comprising: the first judge module the 51, the firstth, and the no execution module 53 of execution module 52, first, the second judge module the 54, the secondth, the no execution module 56 of execution module 55, second, the 3rd judge module the 57, the 3rd are execution modules 58.
Described the first judge module 51 is connected with described received text formatted output module 4, in order to judge whether the positional information in described destination file belongs to region mode; Described first is that execution module 52 is connected with described the first judge module 51, while belonging to region mode in order to the positional information in described destination file, described positional information is referred to region mode; Described the first no execution module 53 is connected with described the first judge module 51, carries out one second judge module while not belonging to region mode in order to the positional information in described destination file; Described the second judge module 54 is connected with described the first no execution module, in order to continue judging whether described positional information belongs to line pattern, alignment pattern or intersecting lens pattern; Described second is that execution module 55 is connected with described the second judge module 54, described positional information is referred to corresponding line pattern, alignment pattern or intersecting lens pattern when belonging to line pattern, alignment pattern or intersecting lens pattern in described positional information; Described the second no execution module 56 is connected with described the second judge module 54, carries out one the 3rd judge module 57 when not belonging to line pattern, alignment pattern or intersecting lens pattern in described positional information; Described the 3rd judge module 57 is connected with described the second no execution module 56, in order to continue judging whether described positional information belongs to single-point pattern or two point pattern; The described the 3rd is that execution module 58 is connected with described the 3rd judge module 57, described positional information is referred to corresponding single-point pattern or two point pattern when belonging to single-point pattern or two point pattern in described positional information.
Further again, the another kind of implementation structure of described automatic clustering module as shown in Figure 8, comprising: order module 51A, cluster analysis module 52A; Described order module 51A is connected with described received text formatted output module 4, in order to be presented as the feature of face, row, column, point according to described fail message, positional information in described destination file is pressed to line direction sequence, then sequence in column direction, the positional information after sequence obtained; Described cluster analysis module 52A is connected with described order module 51A, in order to the positional information after described sequence is carried out to cluster analysis, and the failure mode that finds each positional information to belong to.
Described display module 6 is sorted out result in order to show robotization.
The fail message automatic classification system of CIS imageing sensor of the present invention is realized by software, all modules in this automatic classification system is completely corresponding consistent with the step in automatic classification method in the present invention, the all modules in this automatic classification system is functional module, and the automatic classification system being limited by each functional module is the functional module framework that the computer program of the fail message automatic classification method of realizing CIS imageing sensor recorded by the present invention is realized.
The present invention can accurately understand the image information of CIS imageing sensor output, can locate accurately fail bit, for slip-stick artist FA section provides Data support.The present invention can also carry out mechanized classification statistics fast and effectively by understanding result, facilitate slip-stick artist to failure mode define, statistical query, failure characteristics inspection location.Therefore the processing that the present invention is the relevant BMP picture file of semiconductor test output provides precisely effective method.
In sum, the present invention has effectively overcome various shortcoming of the prior art and tool high industrial utilization.
Above-described embodiment is illustrative principle of the present invention and effect thereof only, but not for limiting the present invention.Any person skilled in the art scholar all can, under spirit of the present invention and category, modify or change above-described embodiment.Therefore, such as in affiliated technical field, have and conventionally know that the knowledgeable, not departing from all equivalence modifications that complete under disclosed spirit and technological thought or changing, must be contained by claim of the present invention.
Claims (12)
1. a fail message automatic classification method for CIS imageing sensor, is characterized in that, the fail message automatic classification method of described CIS imageing sensor comprises:
CIS imageing sensor output detections result picture file; Described testing result picture file is BMP picture file;
Understand described testing result picture file, extract fail message, detailed process comprises: convert described BMP picture file to hex file; The feature that has saltus step according to the pixel color of normal position and fail bit, finds the gray-scale value scope of fail bit by hierarchical cluster analysis method;
Described fail message is converted to the destination file of received text form;
According to self-defined failure mode, the fail message in described destination file is carried out to robotization classification;
Show robotization classification result.
2. the fail message automatic classification method of CIS imageing sensor according to claim 1, it is characterized in that, before the described gray-scale value scope step that finds fail bit by hierarchical cluster analysis method, first described hex file is carried out to frame filtration treatment, the detailed process of filtration treatment is: reject full line in described hex file or permutation black picture element and surpass 90% frame.
3. the fail message automatic classification method of CIS imageing sensor according to claim 1, is characterized in that, the specific implementation process that described fail message is converted to the destination file of received text form comprises:
Read the positional information of the gray-scale value scope that meets described fail bit, described positional information is output into the destination file of received text form.
4. the fail message automatic classification method of CIS imageing sensor according to claim 3, is characterized in that, described self-defined failure mode comprises:
With shape definition failure mode, comprise single-point pattern, two point pattern, line pattern, alignment pattern, intersecting lens pattern, region mode.
5. the fail message automatic classification method of CIS imageing sensor according to claim 4, is characterized in that, the mode that described robotization is sorted out comprises:
According to principle from big to small, by the order of face, line, point, to classify successively, specific implementation process is:
Judge whether the positional information in described destination file belongs to region mode, if be referred to region mode;
Otherwise continue to judge whether described positional information belongs to line pattern, alignment pattern or intersecting lens pattern, if be referred to corresponding line pattern, alignment pattern or intersecting lens pattern;
Otherwise continue to judge whether described positional information belongs to single-point pattern or two point pattern, if be referred to corresponding single-point pattern or two point pattern.
6. according to the fail message automatic classification method of the CIS imageing sensor described in claim 1 or 4, it is characterized in that, the mode that described robotization is sorted out comprises:
According to described fail message, be presented as the feature of face, row, column, point, positional information in described destination file is pressed to line direction sequence, sequence in column direction, then carries out cluster analysis, the failure mode that finds each positional information to belong to by the positional information after sequence again.
7. the fail message automatic classification method of CIS imageing sensor according to claim 1, it is characterized in that, the implementation procedure that result is sorted out in described demonstration robotization comprises: adopt failure mode interface or/and statistics figure surface and interface or/and sensor failure arrangement information interface or/and failure mode and bitmap display interface show that described robotization sorts out result.
8. a fail message automatic classification system for CIS imageing sensor, is characterized in that, the fail message automatic classification system of described CIS imageing sensor comprises:
File layout modular converter, in order to receive the testing result picture file of CIS imageing sensor output, and converts described testing result picture file to hex file;
Fail message extraction module, is connected with described file layout modular converter, in order to there to be the feature of saltus step according to the pixel color of normal position and fail bit, finds the gray-scale value scope of fail bit by hierarchical cluster analysis method;
Received text formatted output module, is connected with described fail message extraction module, in order to described fail message is converted to the destination file of received text form;
Automatic clustering module, is connected with described received text formatted output module, in order to the fail message in described destination file is carried out to robotization classification according to self-defining failure mode.
9. the fail message automatic classification system of CIS imageing sensor according to claim 8, is characterized in that, the fail message automatic classification system of described CIS imageing sensor also comprises:
Filtration module, is connected respectively with described fail message extraction module with described file layout modular converter, in order to full line in described hex file or permutation black picture element are surpassed to 90% frame filtering;
Display module, is connected with described automatic clustering module, in order to show robotization classification result.
10. the fail message automatic classification system of CIS imageing sensor according to claim 8, is characterized in that, described self-defined failure mode comprises:
With shape definition failure mode, comprise single-point pattern, two point pattern, line pattern, alignment pattern, intersecting lens pattern, region mode.
The fail message automatic classification system of 11. CIS imageing sensors according to claim 10, is characterized in that, described automatic clustering module comprises:
The first judge module, is connected with described received text formatted output module, in order to judge whether the positional information in described destination file belongs to region mode;
The firstth, execution module, is connected with described the first judge module, while belonging to region mode in order to the positional information in described destination file, described positional information is referred to region mode;
The first no execution module, is connected with described the first judge module, carries out one second judge module while not belonging to region mode in order to the positional information in described destination file;
The second judge module, is connected with described the first no execution module, in order to continue judging whether described positional information belongs to line pattern, alignment pattern or intersecting lens pattern;
The secondth, execution module, is connected with described the second judge module, described positional information is referred to corresponding line pattern, alignment pattern or intersecting lens pattern when belonging to line pattern, alignment pattern or intersecting lens pattern in described positional information;
The second no execution module, is connected with described the second judge module, carries out one the 3rd judge module when not belonging to line pattern, alignment pattern or intersecting lens pattern in described positional information;
The 3rd judge module, is connected with described the second no execution module, in order to continue judging whether described positional information belongs to single-point pattern or two point pattern;
The 3rd is execution module, is connected with described the 3rd judge module, described positional information is referred to corresponding single-point pattern or two point pattern when belonging to single-point pattern or two point pattern in described positional information.
The fail message automatic classification system of 12. CIS imageing sensors according to claim 10, is characterized in that, described automatic clustering module comprises:
Order module, be connected with described received text formatted output module, in order to be presented as the feature of face, row, column, point according to described fail message, the positional information in described destination file pressed to line direction sequence, sequence in column direction, obtains the positional information after sequence again;
Cluster analysis module, is connected with described order module, in order to the positional information after described sequence is carried out to cluster analysis, and the failure mode that finds each positional information to belong to.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310015074.8A CN103927312B (en) | 2013-01-15 | 2013-01-15 | The fail message automatic classification method and system of a kind of CIS imaging sensors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310015074.8A CN103927312B (en) | 2013-01-15 | 2013-01-15 | The fail message automatic classification method and system of a kind of CIS imaging sensors |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103927312A true CN103927312A (en) | 2014-07-16 |
| CN103927312B CN103927312B (en) | 2017-10-31 |
Family
ID=51145534
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310015074.8A Active CN103927312B (en) | 2013-01-15 | 2013-01-15 | The fail message automatic classification method and system of a kind of CIS imaging sensors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103927312B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105468531A (en) * | 2015-12-25 | 2016-04-06 | 北京金山安全软件有限公司 | Vulnerability mining method and device and electronic equipment |
| CN109117870A (en) * | 2018-07-20 | 2019-01-01 | 深圳大学 | Line cluster extraction method and system based on ArcGIS secondary development |
| CN110019633A (en) * | 2018-07-20 | 2019-07-16 | 深圳大学 | Line density statistical method and system based on ArcGIS secondary development |
| CN110957231A (en) * | 2018-09-26 | 2020-04-03 | 长鑫存储技术有限公司 | Electrical failure pattern discrimination device and discrimination method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001023400A (en) * | 1999-07-07 | 2001-01-26 | Hitachi Ltd | Semiconductor device and its defect analyzing method |
| CN101211947A (en) * | 2006-12-27 | 2008-07-02 | 东部高科股份有限公司 | CMOS image sensor and method for manufacturing the same |
| CN101908382A (en) * | 2009-06-04 | 2010-12-08 | 中芯国际集成电路制造(上海)有限公司 | Data classification analyzing method and device for chip failure |
-
2013
- 2013-01-15 CN CN201310015074.8A patent/CN103927312B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001023400A (en) * | 1999-07-07 | 2001-01-26 | Hitachi Ltd | Semiconductor device and its defect analyzing method |
| CN101211947A (en) * | 2006-12-27 | 2008-07-02 | 东部高科股份有限公司 | CMOS image sensor and method for manufacturing the same |
| CN101908382A (en) * | 2009-06-04 | 2010-12-08 | 中芯国际集成电路制造(上海)有限公司 | Data classification analyzing method and device for chip failure |
Non-Patent Citations (1)
| Title |
|---|
| 徐大诚 等: "基于EZ一USBFxZ 的CMOS数字图像传感器摄像头性能检测系统", 《仪表技术与传感器》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105468531A (en) * | 2015-12-25 | 2016-04-06 | 北京金山安全软件有限公司 | Vulnerability mining method and device and electronic equipment |
| CN105468531B (en) * | 2015-12-25 | 2018-07-31 | 北京金山安全软件有限公司 | Vulnerability mining method and device and electronic equipment |
| CN109117870A (en) * | 2018-07-20 | 2019-01-01 | 深圳大学 | Line cluster extraction method and system based on ArcGIS secondary development |
| CN110019633A (en) * | 2018-07-20 | 2019-07-16 | 深圳大学 | Line density statistical method and system based on ArcGIS secondary development |
| CN109117870B (en) * | 2018-07-20 | 2020-11-13 | 深圳大学 | ArcGIS secondary development-based linear clustering extraction method and system |
| CN110019633B (en) * | 2018-07-20 | 2021-08-10 | 深圳大学 | Line density statistical method and system based on ArcGIS secondary development |
| CN110957231A (en) * | 2018-09-26 | 2020-04-03 | 长鑫存储技术有限公司 | Electrical failure pattern discrimination device and discrimination method |
| CN110957231B (en) * | 2018-09-26 | 2022-03-11 | 长鑫存储技术有限公司 | Electrical failure pattern discrimination device and discrimination method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103927312B (en) | 2017-10-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103577817B (en) | Form recognition method and apparatus | |
| CN109447977B (en) | Visual defect detection method based on multispectral deep convolutional neural network | |
| CN105975929A (en) | Fast pedestrian detection method based on aggregated channel features | |
| CN102254159B (en) | Interpretation method for digital readout instrument | |
| CN102608107A (en) | Automatic test piece interpretation method and system | |
| CN104182750A (en) | Extremum connected domain based Chinese character detection method in natural scene image | |
| CN106295502A (en) | A kind of method for detecting human face and device | |
| CN111047655A (en) | High-definition camera cloth defect detection method based on convolutional neural network | |
| CN103927312A (en) | Automatic classification method and system for failure information of CIS (contact image sensor) | |
| CN103258217A (en) | Pedestrian detection method based on incremental learning | |
| CN103440035A (en) | Gesture recognition system in three-dimensional space and recognition method thereof | |
| CN111667476B (en) | Cloth flaw detection method and device, electronic equipment and readable storage medium | |
| CN111652846B (en) | Semiconductor defect identification method based on characteristic pyramid convolution neural network | |
| CN112014413A (en) | Mobile phone glass cover plate window area defect detection method based on machine vision | |
| CN103245666B (en) | Automatic detecting method for appearance defects of storage battery polar plate | |
| CN102931116B (en) | Synchronous defect detecting method for memorizer | |
| CN110533660B (en) | Method for detecting silk-screen defect of electronic product shell | |
| CN111783608A (en) | Face changing video detection method | |
| CN102109753A (en) | Automatic lens detecting method | |
| CN105509657A (en) | High-precision measurement method of scratch area and grayscale parameter of IC card chip | |
| CN112529901B (en) | Crack identification method in complex environment | |
| CN114428110A (en) | Method and system for detecting defects of fluorescent magnetic powder inspection image of bearing ring | |
| CN114155241A (en) | Foreign matter detection method and device and electronic equipment | |
| CN113781396A (en) | Screen defect detection method, device, equipment and storage medium | |
| CN110458203B (en) | Advertisement image material detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190403 Address after: 518118 No. 18 Gaoxin Road, Longtian Street Export Processing Zone, Pingshan District, Shenzhen City, Guangdong Province Co-patentee after: Semiconductor Manufacturing International (Shanghai) Corporation Patentee after: Zhongxin International Integrated Circuit Manufacturing (Shenzhen) Co., Ltd. Address before: No. 18 Zhangjiang Road, Pudong New Area, Shanghai Patentee before: Semiconductor Manufacturing International (Shanghai) Corporation |