CN108846865A - Graphite crystal localization method in Cast Iron Surface digitized video - Google Patents
Graphite crystal localization method in Cast Iron Surface digitized video Download PDFInfo
- Publication number
- CN108846865A CN108846865A CN201810603174.5A CN201810603174A CN108846865A CN 108846865 A CN108846865 A CN 108846865A CN 201810603174 A CN201810603174 A CN 201810603174A CN 108846865 A CN108846865 A CN 108846865A
- Authority
- CN
- China
- Prior art keywords
- subregion
- threshold
- digitized video
- variance
- cast iron
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
Abstract
The present invention relates to a kind of localization methods of graphite crystal in Cast Iron Surface digitized video, belong to feature location field.After obtaining cast iron process surface number image, this method carries out the division of subregion to digitized video first, is divided into M × N number of subregion of the same size；Secondly the local shape factor factor (local variance and average gradient) and Shannon entropy for calculating each subregion, calculate the relative threshold that resulting entropy determines entropy according to each region；Subregion variance threshold values and average Grads threshold are then determined according to the relative threshold of entropy；The judgement that flake graphite is finally carried out according to subregion variance threshold values and average Grads threshold, then thinks the subregion there are flake graphite if more than subregion variance threshold values or average gradient threshold value and is marked.This method hardware cost is lower, and algorithm robustness is higher, is appropriate for the telltale mark of surface defect or feature.
Description
Technical field
The invention belongs to graphite crystals in image feature positioning field, in particular to a kind of Cast Iron Surface digitized video
Localization method.
Background technique
Cast iron is a kind of iron-carbon alloy, and phosphorus content is higher, is greater than 2.11% with ordinary circumstance.Also contain in cast iron simultaneously
There are many impurity such as more silicon, manganese, phosphorus, sulphur.Wherein existence form of the carbon in cast iron is generally carbide or free
The graphite of state.Flake graphite is formed in process of setting, controls the mechanical performance of gray cast iron.Carrying out surface roughness
When measurement, for contact measurement method since probe is directly contacted with measured surface, flake graphite has no effect on its survey
Measure result；However since graphite reflectivity is lower, it will appear apparent brightness change when carrying out optical measurement, this is to optics
Measurement causes great puzzlement.
It is therefore desirable to develop a kind of side efficiently and accurately positioned to graphite crystal in Cast Iron Surface digitized video
Method.
Summary of the invention
The purpose of the invention is to overcome shortcoming and defect of the existing technology, and provide a kind of Cast Iron Surface number
Graphite crystal localization method in word image, this method hardware cost is lower, and algorithm robustness is higher, is appropriate for surface defect
Or the telltale mark of feature.
To achieve the above object, the technical scheme is that this method includes：
S1:To Cast Iron Surface digitized video carry out subregion division, and use subregion in statistic information searching its
The feature extraction factor；
S2:The local shape factor factor and Shannon entropy of each subregion are calculated, which is part
Variance and average gradient calculate the relative threshold that resulting entropy determines entropy according to each region；
S3:Subregion variance threshold values and average Grads threshold are determined according to the relative threshold of entropy；
S4:The location determination of graphite crystal is carried out, according to subregion variance threshold values and average Grads threshold if more than sub-district
Domain variance threshold values or average gradient threshold value then think the subregion there are graphite crystal and are marked.
Further setting is the calculating all subregion local shape factor factor, local shape factor in the step S2
The factor is defined as：
Wherein var indicates the local variance of digitized video subregion, and the size of subregion is 32 × 32；Grin indicates number
The local mean gradient of word image subregion；gi,jIndicate the pixel grey scale information of digitized video, wherein i, j indicates pixel position
It sets, m, n indicates digitized video sub-window position.
Further setting is that the relative threshold calculating that Shannon entropy carries out characteristic image in the step S2 specifically includes：
The definition of Shannon entropy is：
The relative threshold of its entropy can be identified as：
Wherein,Indicate the Shannon entropy average value of all subregions.
Further setting is that the step S4 is specifically included：
Local shape factor factor dynamic threshold can be set, and specific algorithm is：
Wherein varthIndicate subregion variance threshold values,Indicate subregion mean of variance；grinthIndicate subregion ladder
Threshold value is spent,Indicate subregion gradient average value, λvFor variance weight coefficient, λgFor gradient weight coefficient.
It is flake graphite that further setting, which is the graphite crystal,.
The technical solution of the application using entropy, three kinds of statistics of variance and gradient to Cast Iron Surface graphite crystal position into
Row positioning, achieves extraordinary effect.This method hardware cost is lower, and algorithm robustness is higher, is appropriate for surface defect
Or the telltale mark of feature.Specific effect is shown in specification embodiment.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will to embodiment or
Attached drawing needed to be used in the description of the prior art is briefly described, it should be apparent that, the accompanying drawings in the following description is only
Some embodiments of the present invention, for those of ordinary skill in the art, without any creative labor,
It obtains other drawings based on these drawings and still falls within scope of the invention.
Fig. 1 flake graphite location algorithm flow chart；
Fig. 2 coaxial light source layout drawing；
Fig. 3 cast iron process surface platelets graphite metallographic microscope；
The distribution of Fig. 4 variance and its threshold figure；
Fig. 5 gradient distribution and its threshold figure；
Fig. 6 the method for the present invention flow chart.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, the present invention is made into one below in conjunction with attached drawing
Step ground detailed description.
Technical solution of the present invention shoots cast iron process surface using coaxial light source shown in Fig. 2, number
Image destroys the continuity of matrix as shown in figure 3, since graphite flake has dissection to matrix, so that between the two
Boundary is not very clear and legible, is accurately positioned using traditional morphological method is more difficult to it.
1) judgement that flake graphite is directly carried out to single pixel point is complex, needs to carry out it in certain region
Division forms several sub-blocks and then uses its feature extraction factor of statistic information searching in subregion.Therefore, subregion
Size determines the precision of flake graphite location algorithm.
In the specific use process, it is too sensitive and navigate to casting to will lead to statistic information for too small subregion size
The normal region of iron；And then to will lead to flake graphite location algorithm precision poor for too big subregion.Therefore selection is sub herein
The size in region is 8 × 8 pixels.
2) gray scale of the body surface digitized video of single material is tended to be uniformly distributed.Therefore, the feature extraction factor
Major function is the special digitized video intensity profile subregion of identification.In this application, two different local shape factors
Operator (local variance and average gradient) is used for the positioning of flake graphite.
3) histogram of digitized video characterizes the probability distribution of its grayscale information.And the comentropy of digitized video then can be with
For carrying out average information of the aggregation properties of digitized video intensity profile to describe digital image signal source.Therefore originally
Application determines that the relative threshold of entropy can be identified as using the dynamic threshold that Shannon entropy carries out characteristic image：
Wherein,Indicate the Shannon entropy average value of all subregions.
4) by being repeatedly adjusted experiment, local variance and average ladder to local shape factor operator weight coefficient
The dynamic threshold of degree can be arranged to：
Wherein varthIndicate subregion variance threshold values,Indicate subregion mean of variance；grinthIndicate subregion
Grads threshold,Indicate subregion gradient average value.
5) to Cast Iron Surface shown in Fig. 3 carry out flake graphite detection and localization, variance distribution and its threshold value as shown in figure 4,
Wherein blue color planes are its threshold value position, in upper figure, sub-window position corresponding higher than the grid point of threshold plane
It is considered the position containing flake graphite.It marks these regions and zero setting carries out the interpolation operation of next step.Comparison diagram 3 and 4 can
Know, using only variance, there is no the positions for completely finding all flake graphites.Since variance only measures digitized video grayscale information
Dispersion degree, therefore it is only sensitive to the high frequency section of digital image greyscale variation, and in some flake graphite regions,
Grayscale information variation be not very greatly, the state in the low variance of low ash degree therefore needs using other statistics carry out into
The feature extraction of one step.
Flake graphite detection and localization, gradient distribution and threshold plane are carried out to Cast Iron Surface digitized video using gradient
As shown in figure 5, wherein green color plane is its threshold plane position.It is distributed compared with variance, gradient is shown with variance more
Strong flake graphite stationkeeping ability, this is because gradient is the derivation of digitized video two-dimensional discrete function, to grayscale information
Change very sensitive, therefore in contrast stronger flake graphite detection and localization ability can be obtained.
To variance and gradient to Cast Iron Surface digitized video carry out the subinterval position of flake graphite detection and localization acquisition into
Row union operation, the region of acquisition are the subinterval position containing flake graphite that this algorithm obtains.
Those of ordinary skill in the art will appreciate that implement the method for the above embodiments be can be with
Relevant hardware is instructed to complete by program, the program can be stored in a computer readable storage medium,
The storage medium, such as ROM/RAM, disk, CD.
The above disclosure is only the preferred embodiments of the present invention, cannot limit the right of the present invention with this certainly
Range, therefore equivalent changes made in accordance with the claims of the present invention, are still within the scope of the present invention.
Claims (5)
1. graphite crystal localization method in a kind of Cast Iron Surface digitized video, it is characterised in that this method includes：
S1:The division of subregion is carried out to Cast Iron Surface digitized video, and uses its feature of statistic information searching in subregion
Extraction factor；
S2:The local shape factor factor and Shannon entropy of each subregion are calculated, which is local variance
And average gradient, the relative threshold that resulting entropy determines entropy is calculated according to each region；
S3:Subregion variance threshold values and average Grads threshold are determined according to the relative threshold of entropy；
S4:The location determination of graphite crystal is carried out, according to subregion variance threshold values and average Grads threshold if more than subregion side
Poor threshold value or average gradient threshold value then think the subregion there are graphite crystal and are marked.
2. graphite crystal localization method in a kind of Cast Iron Surface digitized video according to claim 1, it is characterised in that institute
The all subregion local shape factor factor is calculated in the step S2 stated, the local shape factor factor is defined as：
Wherein var indicates the local variance of digitized video subregion, and the size of subregion is 32 × 32；Grin indicates digitized video
The local mean gradient of subregion；gi,jIndicate the pixel grey scale information of digitized video, wherein i, j indicates location of pixels, m, n table
Show digitized video sub-window position.
3. graphite crystal localization method in a kind of Cast Iron Surface digitized video according to claim 2, it is characterised in that：Institute
The relative threshold calculating that Shannon entropy carries out characteristic image in the step S2 stated specifically includes：The definition of Shannon entropy is：
The relative threshold of its entropy can be identified as：
Wherein,Indicate the Shannon entropy average value of all subregions.
4. graphite crystal localization method in a kind of Cast Iron Surface digitized video according to claim 3, it is characterised in that：Institute
The step S4 stated is specifically included：
Local shape factor factor dynamic threshold can be set, and specific algorithm is：
Wherein varthIndicate subregion variance threshold values,Indicate subregion mean of variance；grinthIndicate subregion gradient threshold
Value,Indicate subregion gradient average value, λvFor variance weight coefficient, λgFor gradient weight coefficient.
5. graphite crystal localization method in Cast Iron Surface digitized video according to claim 1, it is characterised in that：Described
Graphite crystal is flake graphite.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810603174.5A CN108846865A (en) | 2018-06-12 | 2018-06-12 | Graphite crystal localization method in Cast Iron Surface digitized video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810603174.5A CN108846865A (en) | 2018-06-12 | 2018-06-12 | Graphite crystal localization method in Cast Iron Surface digitized video |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108846865A true CN108846865A (en) | 2018-11-20 |
Family
ID=64211086
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810603174.5A Withdrawn CN108846865A (en) | 2018-06-12 | 2018-06-12 | Graphite crystal localization method in Cast Iron Surface digitized video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108846865A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706951A (en) * | 2009-11-20 | 2010-05-12 | 上海电机学院 | Method, device and system for objectively evaluating pneumatic optical image quality based on feature fusion |
| CN103218619A (en) * | 2013-03-15 | 2013-07-24 | 华南理工大学 | Image aesthetics evaluating method |
-
2018
- 2018-06-12 CN CN201810603174.5A patent/CN108846865A/en not_active Withdrawn
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101706951A (en) * | 2009-11-20 | 2010-05-12 | 上海电机学院 | Method, device and system for objectively evaluating pneumatic optical image quality based on feature fusion |
| CN103218619A (en) * | 2013-03-15 | 2013-07-24 | 华南理工大学 | Image aesthetics evaluating method |
Non-Patent Citations (1)
| Title |
|---|
| QIWU LUO 等: "A cost-effective and automatic surface defect inspection system for hot-rolled flat steel", 《ROBOTICSANDCOMPUTER-INTEGRATEDMANUFACTURING》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108629775B (en) | Thermal state high-speed wire rod surface image processing method | |
| CN110288561A (en) | Refractory brick surface scratch recognition methods based on frequency filtering enhancing | |
| CN115100221B (en) | Glass defect segmentation method | |
| CN116309537B (en) | Defect detection method for oil stain on surface of tab die | |
| CN110286124A (en) | Refractory brick measuring system based on machine vision | |
| CN104655403A (en) | Luminance uniformity test method of dot-matrix light source | |
| CN106530271A (en) | Infrared image significance detection method | |
| CN116542969B (en) | Road asphalt adhesion detection method based on vision technology | |
| CN112488928A (en) | Transformer oil conservator oil level online monitoring device and method based on image recognition technology | |
| CN112595726A (en) | Pixel defect detection method of OLED (organic light emitting diode) micro display device | |
| CN115311280B (en) | Defect detection method for silicon carbide wafer | |
| CN113781446A (en) | Method and device for detecting greasy dirt on glass fiber surface, storage medium and electronic equipment | |
| Rana et al. | An evaluation of HDR image matching under extreme illumination changes | |
| TWI767642B (en) | Wafer inspection method and apparatus thereof | |
| CN111311580A (en) | Steam drum liquid level abnormity identification method and system based on image identification | |
| CN108846865A (en) | Graphite crystal localization method in Cast Iron Surface digitized video | |
| CN116309557B (en) | Method for detecting fracture of track shoe of excavator | |
| Domínguez-Nicolas et al. | Algorithm for automatic detection and measurement of Vickers indentation hardness using image processing | |
| JP4317865B2 (en) | Corrosion evaluation method for ferrous structures | |
| CN108804997A (en) | Visual angle scene image processing platform | |
| CN108280825A (en) | A kind of liquid crystal display emargintion detection method | |
| CN115049621A (en) | Micropipe defect detection method, device, equipment, storage medium and program product | |
| CN116563298A (en) | Cross line center sub-pixel detection method based on Gaussian fitting | |
| Fedotkin et al. | Automatic processing of microhardness images using computer vision methods | |
| CN108734706B (en) | Rotor winding image detection method fusing regional distribution characteristics and edge scale angle information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181120 |