CN108052950A - A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA - Google Patents
A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA Download PDFInfo
- Publication number
- CN108052950A CN108052950A CN201711293082.3A CN201711293082A CN108052950A CN 108052950 A CN108052950 A CN 108052950A CN 201711293082 A CN201711293082 A CN 201711293082A CN 108052950 A CN108052950 A CN 108052950A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- flame
- image
- block diagram
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA of the present invention, including：The video image of flame is gathered, RGB image is converted into two-dimensional matrix；Dimensionality reduction is carried out to two-dimensional matrix using PCA methods；By the matrix normalization after dimensionality reduction between [0,255], score block diagram is obtained；The score block diagram of different operating modes is compared, finds out and changes apparent region in figure and processing is marked, labeled area maps are returned into original RGB image, obtain flame segmentation figure picture；5 kinds of flame luminance area size, flame color species number, flame region color average, entire image color average and flame brightness value characteristics are calculated respectively by marked area in score block diagram.This method can effectively split flame region, and segmentation effect is good, and 5 kinds of characteristics are calculated by the image to segmentation, and result is applied on producing condition classification model, obtain higher classification accuracy.
Description
Technical field
The invention belongs to pattern-recognition and field of artificial intelligence, a kind of electric melting magnesium furnace based on MIA especially set out moves
State flame splits and feature extracting method.
Background technology
Fused magnesite has high purity, fusing point height, insulation performance strong and the characteristic of compact structure, in chemical industry, building, family
A variety of industries such as electricity, metallurgy, military affairs and field are widely used, and are good fireproof raw materials.
When producing fused magnesite using arc melting, corresponding melting operating mode, charging operating mode, exhaust operating mode, underburnt operating mode
The brightness of fire door flame, color, metamorphosis it is different.Wherein, under melting operating mode, flame brightness is moderate, flame color
Compare abundant, flame forms variation is general；Feed operating mode under, flame brightness ratio is dark, and flame color is general, metamorphosis compared with
Slowly；It is vented under operating mode, flame brightness is brighter, and flame color is relatively simple, and flame forms variation is very fast.So fire door flame
Situation of change is an important evidence of work condition judging.
At present, the information contained in fire door flame needs the mode of manual inspection, and production line is gone to pass through " seeing fire "
Experience obtains.But manual inspection has the following problems：1) experience of accuracy and the operating personnel judged and state is related,
Easy missing inspection, flase drop；2) in-situ production environment is severe (strong light, high temperature, dust etc.), and labor intensity is big, dangerous high, is unsuitable for
Worker's live inspection for a long time；3) fire door flame and surrounding smog dust obscurity boundary, " seeing fire " process are easily disturbed.So
Enterprise needs one kind that can split fire door flame region from image, and with specific number represent flame color, brightness,
The characteristic informations such as form, and in subsequent work condition judging.
With the application of industrial camera, the precision of video image acquisition gradually steps up, and it is raw to gather industry with industrial camera
Then image in production obtains area-of-interest by image partition method, and calculates the characteristic in the region.The process is
Through being widely used in the industrial productions such as the monitoring of natural gas fuel oil quality, rotary kiln combustion position.But due to fire
Flame is dynamic, without the form of fixed rule, for different industrial backgrounds, change frequency, color, the brightness of flame etc.
Feature is also very doubt, and in some independent industrial process, dynamic flame image is all existing in a variety of forms, and
It is tightly combined with background environment, so being difficult to adopt the thought of threshold method, image segmentation algorithm is made to be suitable for all operating modes.
The content of the invention
The embodiment of the present invention provides a kind of electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, can
Dynamic flame region is effectively split, by calculating 5 kinds of characteristics to the image of segmentation, and after result is applied to
On continuous producing condition classification model, higher classification accuracy is obtained.
The present invention provides a kind of electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, including following step
Suddenly：
Step 1：The video image of flame is gathered, and the RGB image of single frames is converted into two-dimensional matrix；
Step 2：Dimension-reduction treatment carries out the two-dimensional matrix using Principal Component Analysis Method, and will by image reconstruction technique
The pivot of selection maps back RGB image space and is verified, to ensure that the image after dimensionality reduction can replace original image；
Step 3：By the matrix normalization with two column vectors after dimensionality reduction between [0,255], then with every in matrix
Two of row data, which are worth, as the location information under XY coordinate systems, counts number of pixels of the matrix with same coordinate value, from
And obtain score block diagram；
Step 4：Step 1-3 is repeated for the RGB image of the single frames of different operating modes, by the score column of the different operating modes of acquisition
Shape figure is compared, and is found out in the score block diagram of different operating modes and is changed apparent region, to changing apparent region into rower
Note processing, and the area maps being labeled in the score block diagram are returned into original RGB image, to obtain flame segmentation figure picture；
Step 5：The region being labeled in score block diagram is finely adjusted, with the definite marked region of acquisition and accurately
Flame segmentation figure picture, and acquisition is split to single frames RGB image all in video image often according to definite marked region
The flame segmentation figure picture of frame；
Step 6：By feature extraction formula, flame luminance area is calculated respectively by marked area in score block diagram
5 kinds of size, flame color species number, flame region color average, entire image color average and flame brightness value spies
Levy data.
A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA of the present invention, utilizes industrial camera
Electric-melting magnesium production scene procedural image is obtained, the extraneous areas in image from flame farther out, Ran Houyun are rejected using stationary window
With multivariate image analysis (MIA) method, flame region interested from background is split, and calculates relative token number
According to.This method can effectively split fire door flame region, and segmentation effect is than the segmentation effect based on threshold method
It is good very much, 5 kinds of characteristics are calculated by the image to segmentation, and result is applied on subsequent producing condition classification model, it obtains
Obtain higher classification accuracy.
Description of the drawings
Fig. 1 is the flow chart of a kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA of the present invention；
Fig. 2 is the flame image of the first operating mode of the present invention；
Fig. 3 is the flame image of second of operating mode of the present invention；
Fig. 4 is the flame image of the third operating mode of the present invention；
Fig. 5 is the flame image of the 4th kind of operating mode of the present invention；
Fig. 6 is the score block diagram of the first operating mode of the present invention；
Fig. 7 is the score block diagram of second of operating mode of the present invention；
Fig. 8 is the score block diagram of the third operating mode of the present invention；
Fig. 9 is the score block diagram of the 4th kind of operating mode of the present invention；
Figure 10 is the schematic diagram that is labeled of definite marked region in score block diagram；
Figure 11 is the schematic diagram being labeled in the enclosure region of the definite marked region of score block diagram；
Figure 12 is to map back the signal of original RGB image using being labeled for the marked region determined in score block diagram
Figure；
Figure 13 is to be labeled the signal for mapping back original RGB image using the marked region of definite marked region attachment
Figure；
A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA that Figure 14 is the present invention split one group
The flame segmentation figure picture that image obtains；
Figure 15 is extracted for a kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA of the present invention
5 kinds of characteristic trend charts.
Specific embodiment
Set forth herein a kind of electricity for being based on multivariate image analysis method (Multivariate Image Analysis, MIA)
Molten magnesium stove fire door dynamic flame image segmentation and feature extracting method.Electric-melting magnesium production scene procedure chart is obtained using industrial camera
Picture rejects the extraneous areas in image from flame farther out using stationary window, then with multivariate image analysis (MIA) side
Method splits flame region interested from background, and calculates correlated characteristic data.
A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA of the present invention includes as shown in Figure 1
Following steps：
Step 1：The video image of flame is gathered, and the RGB image of single frames is converted into two-dimensional matrix；Step 1 is specifically wrapped
It includes：
Step 1.1：Window segmentation is fixed in the video image that industrial camera is collected, and rejects flame peripheral region pair
Interference caused by image is split, while computer operating cost is reduced, improve single frames graphics process speed；
Step 1.2：The matrix being transformed to single frames flame RGB image by matrixing under two-dimensional space.
Image after stationary window is split has 3 passages in rgb space, i.e. dimension is m × n × 3, will each be led to
Road stretched by row, such as R passages, becomes the data of mn rows after stretching, then becomes mn × 3 after three channel extrusions
Matrix, i.e.,
Wherein, I(m,n,3)Represent original RGB image, I1(m,n,3)It represents I(m,n,3)Expand into the matrix after two dimension.
Step 2：Using Principal Component Analysis Method (principal component analysis, PCA) to the Two-Dimensional Moment
Battle array carries out dimension-reduction treatment, and passes through image reconstruction technique the pivot of selection is mapped back RGB image space and verify, to ensure
Image after dimensionality reduction can replace original image；The step 2 specifically includes：
Step 2.1：The two-dimensional matrix of above-mentioned mn × 3 is carried out by dimensionality reduction by Principal Component Analysis Method, chooses a certain amount of pivot
Matrix after quantitative commitments dimensionality reduction can characterize the information of initial data 99%；
The core algorithm of Principal Component Analysis Method is as follows：
Wherein, PC be pivot number, paRepresent loading vector, taRepresent the matrix after score vector, that is, dimensionality reduction.
Asking for pa=(1 ..., PC), ta, it is necessary to consider the problems of that row dimension is excessive during (a=1 ..., PC),
The image space of one mn=512 × 512 corresponds to 262144 rows, the so many row formed afterwards for Multivariate image expansion
Dimension matrix by a core algorithm to simplify, it is necessary to calculate.The core concept of above-mentioned algorithm is to the image after expansion, head
Equalization first is carried out for each attribute, then builds a nuclear matrix I1 TI1, obtained low-dimensional Matrix C (3,3) is carried out
Singular value decomposition (singular value decomposition, SVD), by feature vector by character pair value size from greatly to
Into matrix, each row of obtained eigenvectors matrix are to load vector p for minispreada=(1 ..., PC).On this basis
According to ta=I1pa, calculate score vector ta(a=1 ..., PC).
Step 2.2：The pivot of selection is mapped back by RGB image space by image reconstruction technique, is compared by observing, really
Original image can be replaced by determining the image after dimensionality reduction.
When it is implemented, the value of pivot can be determined by following formula, it is original more than needs to meet the data after dimensionality reduction
Pictorial information, determine the image after dimensionality reduction can replace original image, wherein PC (PC≤3) be pivot number, λkTo pass through
The characteristic value of covariance matrix is acquired during principal component analysis processing.
Wherein, g represents that the pivot of selection can characterize the percentage of original image information for threshold value, generally takes g >=0.95.
For score vector ta(a=1 ..., PC), t1In the original graphical message that includes it is most, t2Comprising information time at most, with this
Analogize.Original Multivariate image is reconstructed in PC prevailing main compositions before selection, and residual matrix E is ignored,
So as to be eliminated in original image largely structureless noise.
It, can be by equation below, according to score vector t after above-mentioned computing is completedaWith loading vector paTo original graph
New RGB image matrix is obtained as being reconstructed, compressed distorted image condition is observed by Comparative result.
Respectively image reconstruction is carried out using the first two pivot, first pivot, second pivot and the 3rd pivot.To original
After beginning fire door flame image carries out PCA processing, the first two characteristic value is taken to bring calculating into, the results show is more than 99.5%, i.e. basis
The result that the first two pivot is reconstructed includes the information of original image more than 99.5%.Then weight is carried out according to each pivot
Structure, they reconstructed image characterization original image information size as their characteristic value sizes from big to small.Wherein, it is preceding
Two pivot reconstruction results and original image are very close, comprising information it is most.First pivot, second pivot and the 3rd
It is increasingly severe that a pivot corresponds to original image distortion situation.When selecting the first two pivot characterization raw information, then the 3rd
The corresponding numerical value of a pivot is exactly residual matrix E.
Step 3：By the matrix normalization with two column vectors after dimensionality reduction between [0,255], then with every in matrix
Two of row data, which are worth, as the location information under XY coordinate systems, counts number of pixels of the matrix with same coordinate value, from
And obtain score block diagram；
When it is implemented, using following formula by the matrix normalization with two column vectors after dimensionality reduction between [0,255]：
Wherein, siRepresent two column vectors after normalization, tiRepresent two column vectors before normalization, ti,minRepresent column vector
tiThe element of middle minimum, ti,maxRepresent column vector tiThe element of middle maximum.
When it is implemented, score block diagram is obtained using following formula：
Wherein, TTx,yFor the number of similary pixel that coordinate in score block diagram is (x, y), b is certain row of matrix,
s1,bRepresent vector s1The element of b rows, s2,bRepresent vector s2The element of b rows.
Step 4：Step 1-3 is repeated for the RGB image of the single frames of different operating modes, by the score column of the different operating modes of acquisition
Shape figure is compared, and is found out in the score block diagram of different operating modes and is changed apparent region, to changing apparent region into rower
Note processing, and the area maps being labeled in the score block diagram are returned into original RGB image, to obtain flame segmentation figure picture；
It will be verified below by taking the fire door flame in electric melting magnesium furnace production process as an example.
First, according to the experience of operator, we choose 4 kinds of typical magnesium stove fire door flame-shapeds from multitude of video data
State, represents charging, normal smelting respectively, and underburnt is vented 4 kinds of operating modes.As shown in Figures 2 to 5.
Then, repeat step 1-3 and respectively obtain the score block diagram of 4 kinds of operating modes as shown in Figures 6 to 9.Above-mentioned four figures
It is projection of the fire door flame after being handled by MIA under xy coordinate systems respectively, it is s that wherein x-axis is corresponding1, y-axis is corresponding
For s2, they reflect aggregation situation of the fire door flame color luminance information in two-dimensional space, and the brighter region of color represents stove
Number of the mouth with this color is more.
Finally, by comparing the distribution situation of four score block diagram Midst density figures, it has been found that in x-axis [0~50], y
In the region of axis [0~200], the variation of score block diagram is the most apparent with the variation of picture color and brightness.
Step 5：The region being labeled in score block diagram is finely adjusted, with the definite marked region of acquisition and accurately
Flame segmentation figure picture, and acquisition is split to single frames RGB image all in video image often according to definite marked region
The flame segmentation figure picture of frame；The step 5 includes：
Step 5.1：The region being labeled in score block diagram is finely adjusted, and obtains new flame segmentation figure picture, if
Segmentation effect improves significantly, then retains new marked region, and definite marked region and accurate is obtained by repeatedly adjusting
Flame segmentation figure picture；
Step 5.2：By proof method in described definite one piece of validation region of marked region surrounding markings, by validation region
Original RGB image is mapped back, if the corresponding segmentation figure image position of validation region shows described definite around flame image
Marked region meets the requirement of flame segmentation.
When it is implemented, in order to which the region that verification step 5.1 obtains is interested flame region, we are in score column
Another region is marked on shape figure, the interested region before, while show in original image the color area of mark
Domain.One piece of s1 [0,50] section of mark and the section of s2 [0,200], Figure 12 show mapping on score block diagram shown in Figure 10
Return original image as a result, i.e. segmentation result image, Figure 11 are the areas adjacent marked in Fig. 10, re-flag one piece it is polygon
Shape region maps back original image and obtains the smog image of Figure 13, i.e. flame periphery.If the corresponding segmentation figure image position of validation region in
Around flame image, then show that the definite marked region meets the requirement of flame segmentation.
Step 6：By feature extraction formula, flame luminance area is calculated respectively by marked area in score block diagram
5 kinds of size, flame color species number, flame region color average, entire image color average and flame brightness value spies
Levy data.
Wherein, flame luminance area size refers to the size of the gained area-of-interest after over-segmentation, original RGB image
Each pixel value is projected in score block diagram, so area-of-interest size is mark zone in score block diagram
The sum of all numerical value in domain calculate flame luminance area size according to the following formula：
Wherein, A represents flame luminance area size, TTx,yFor the similary pixel that coordinate in score block diagram is (x, y)
Number, Mi,jMarked region in=1 pair of reserved portion block diagram.
Each coordinate of the numerical value in marked region more than 1 in step 6 in score block diagram corresponds to original RGB figures
A kind of color of picture, flame color species number be score block diagram in marked region in numerical value more than 0 coordinate number and,
Flame color species number is sought according to the following formula：
Wherein, C represents flame color species number, TTx,yFor similary pixel that coordinate in score block diagram is (x, y)
Number, Mi,jMarked region in=1 pair of reserved portion block diagram.
Flame region color average is calculated in step 6 according to the following formula：
Wherein, sfRepresent flame region color average, TTx,yFor the similary pixel that coordinate in score block diagram is (x, y)
The number of point, A represent flame luminance area size；
Entire image color average is calculated according to the following formula：
Wherein, smFor entire image color average, TTx,yFor the similary pixel that coordinate in score block diagram is (x, y)
Number, M × N be entire image size.
Flame brightness value is calculated in step 6 is specially：
First, score block diagram is reduced by RGB image according to following equation：
Wherein, t1,maxRepresent column vector t1The element of middle maximum, t1,minRepresent column vector t1The element of middle minimum, t2,maxGeneration
Table column vector t2The element of middle maximum, t2,minRepresent column vector t2The element of middle minimum, p1To load vector pa=(1 ..., PC)
First row, p2To load vector paThe secondary series of=(1 ..., PC).
Secondly, gray level image is calculated according to the following formula：
Lxy=[R, G, B]xy[0.299,0.587,0.114]T (13)
Finally, flame brightness value is calculated according to the following formula：
Wherein, TTx,yFor the number of similary pixel that coordinate in score block diagram is (x, y).
One section is handled using the segmentation of the electric melting magnesium furnace dynamic flame based on MIA of the present invention and feature extracting method about 58 seconds
Electric melting magnesium furnace production video data, obtained 1500 images will be split and preserved, wherein the region of interest split
Domain part is that the flame image result per frame is as shown in figure 14.With common image classification method compare, using invention based on
The electric melting magnesium furnace dynamic flame segmentation of MIA and feature extracting method realize that image of interest region segmentation result effect is more preferable.
For the region being labeled in the score block diagram of each two field picture, characteristic is carried out using formula (6) (7) (8) (9) (14)
It calculates, the tendency chart for obtaining 5 kinds of flame characteristic data is as shown in figure 15.Corresponding video image and region of interest area image, observation
Curvilinear motion situation finds that this 5 kinds of characteristic variation tendencies have very strong correlation with video image situation of change.Specifically
It is that presentation is periodically variable to be embodied in several operating modes of electric melting magnesium furnace under site environment, and splits the big of the image of interest of gained
Small, color, form are also that periodically variation is presented, and it is in 5 kinds of characteristic trend that this variation, which more intuitively embodies,
The approximately periodic variation of figure.So this 5 kinds of characteristics subsequently using machine learning carry out classify ensure higher essence
Degree has great importance.
The foregoing is merely presently preferred embodiments of the present invention, the thought being not intended to limit the invention, all the present invention's
Within spirit and principle, any modifications, equivalent replacements and improvements are made should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA, which is characterized in that including following step
Suddenly：
Step 1：The video image of flame is gathered, and the RGB image of single frames is converted into two-dimensional matrix；
Step 2：Dimension-reduction treatment is carried out to the two-dimensional matrix using Principal Component Analysis Method, and passes through image reconstruction technique to choose
Pivot map back RGB image space and verified, to ensure that the image after dimensionality reduction can replace original image；
Step 3：By the matrix normalization with two column vectors after dimensionality reduction between [0,255], then in matrix per line number
According to two values as the location information under XY coordinate systems, number of pixels of the matrix with same coordinate value is counted, so as to obtain
Obtain score block diagram；
Step 4：Step 1-3 is repeated for the RGB image of the single frames of different operating modes, by the score block diagram of the different operating modes of acquisition
It is compared, finds out in the score block diagram of different operating modes and change apparent region, place is marked to changing apparent region
Reason, and the area maps being labeled in the score block diagram are returned into original RGB image, to obtain flame segmentation figure picture；
Step 5：The region being labeled in score block diagram is finely adjusted, to obtain definite marked region and accurate flame
Segmentation figure picture, and single frames RGB image all in video image is split according to definite marked region and is obtained per frame
Flame segmentation figure picture；
Step 6：By feature extraction formula, it is big that flame luminance area is calculated respectively by marked area in score block diagram
5 kinds of small, flame color species number, flame region color average, entire image color average and flame brightness value features
Data.
2. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In the step 1 includes：
Step 1.1：Window segmentation is fixed in the video image that industrial camera is collected, and rejects flame peripheral region to image
Interference caused by segmentation；
Step 1.2：The matrix being transformed to single frames flame RGB image by matrixing under two-dimensional space.
3. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In the step 2 includes：
Step 2.1：Above-mentioned two-dimensional matrix is carried out by dimensionality reduction by Principal Component Analysis Method, chooses a certain amount of pivot quantitative commitments drop
Matrix after dimension can characterize the information of initial data 99%；
Step 2.2：The pivot of selection is mapped back by RGB image space by image reconstruction technique, is compared by observing, determines drop
Image after dimension can replace original image.
4. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In, in the step 3 using following formula by the matrix normalization with two column vectors after dimensionality reduction between [0,255]：
<mrow>
<msub>
<mi>s</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mi>R</mi>
<mi>o</mi>
<mi>u</mi>
<mi>n</mi>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<msub>
<mi>t</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>t</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>min</mi>
</mrow>
</msub>
</mrow>
<mrow>
<msub>
<mi>t</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>t</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>min</mi>
</mrow>
</msub>
</mrow>
</mfrac>
<mo>&times;</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, siRepresent two column vectors after normalization, tiRepresent two column vectors before normalization, ti,minRepresent column vector tiIn
Minimum element, ti,maxRepresent column vector tiThe element of middle maximum.
5. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In using following formula acquisition score block diagram in the step 3：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>TT</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<msub>
<mi>&Sigma;</mi>
<mi>b</mi>
</msub>
<mn>1</mn>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mo>&ForAll;</mo>
<mi>b</mi>
<mo>,</mo>
<mi>x</mi>
<mo>=</mo>
<msub>
<mi>s</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>b</mi>
</mrow>
</msub>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<msub>
<mi>s</mi>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>b</mi>
</mrow>
</msub>
<mo>,</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, TTx,yFor the number of similary pixel that coordinate in score block diagram is (x, y), b is certain row of matrix, s1,bTable
Show vectorial s1The element of b rows, s2,bRepresent vector s2The element of b rows.
6. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In the step 5 includes：
Step 5.1：The region being labeled in score block diagram is finely adjusted, and obtains new flame segmentation figure picture, if segmentation
Effect improves significantly, then retains new marked region, and definite marked region and accurate fire are obtained by repeatedly adjusting
Flame segmentation figure picture；
Step 5.2：By proof method in described definite one piece of validation region of marked region surrounding markings, validation region is mapped
Original RGB image is returned, if the corresponding segmentation figure image position of validation region shows the definite mark around flame image
Region meets the requirement of flame segmentation.
7. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In the step 6 Flame luminance area size refers to the size of the gained area-of-interest after over-segmentation, original RGB image
Each pixel value be projected in score block diagram, so area-of-interest size in score block diagram be mark
The sum of all numerical value in region calculate flame luminance area size according to the following formula：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>A</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</munder>
<msub>
<mi>TT</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mo>&ForAll;</mo>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>,</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, A represents flame luminance area size, TTx,yFor of similary pixel that coordinate in score block diagram is (x, y)
Number, Mi,jMarked region in=1 pair of reserved portion block diagram.
8. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In each coordinate of the numerical value in marked region more than 0 in the step 6 in score block diagram corresponds to original RGB figures
A kind of color of picture, flame color species number be score block diagram in marked region in numerical value more than 1 coordinate number and,
Flame color species number is sought according to the following formula：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>C</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msub>
<mi>TT</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>></mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mo>&ForAll;</mo>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>,</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, C represents flame color species number, TTx,yFor the number of similary pixel that coordinate in score block diagram is (x, y),
Mi,jMarked region in=1 pair of reserved portion block diagram.
9. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In flame region color average is calculated in the step 6 according to the following formula：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>s</mi>
<mi>f</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>&Sigma;</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<msub>
<mi>TT</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mi>x</mi>
</mrow>
<mi>A</mi>
</mfrac>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mo>&ForAll;</mo>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>,</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, sfRepresent flame region color average, TTx,yFor similary pixel that coordinate in score block diagram is (x, y)
Number, A represent flame luminance area size；
Entire image color average is calculated according to the following formula：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>s</mi>
<mi>m</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<msub>
<mi>TT</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mi>x</mi>
</mrow>
<mrow>
<mi>m</mi>
<mo>&times;</mo>
<mi>n</mi>
</mrow>
</mfrac>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mo>&ForAll;</mo>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>,</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, smFor entire image color average, TTx,yFor of similary pixel that coordinate in score block diagram is (x, y)
Number, m × n are the size of entire image.
10. the electric melting magnesium furnace dynamic flame segmentation based on MIA and feature extracting method, feature exist as described in claim 1
In flame brightness value is calculated in the step 6 is specially：
First, score block diagram is reduced by RGB image according to following equation：
<mrow>
<msub>
<mrow>
<mo>&lsqb;</mo>
<mi>R</mi>
<mo>,</mo>
<mi>G</mi>
<mo>,</mo>
<mi>B</mi>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<mi>x</mi>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<msub>
<mi>t</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<msubsup>
<mi>p</mi>
<mn>1</mn>
<mi>T</mi>
</msubsup>
<mo>+</mo>
<msub>
<mi>t</mi>
<mn>2</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<msubsup>
<mi>p</mi>
<mn>2</mn>
<mi>T</mi>
</msubsup>
<mo>,</mo>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...255</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>t</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>x</mi>
<mo>&times;</mo>
<mrow>
<mo>(</mo>
<msub>
<mi>t</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>t</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mn>255</mn>
</mfrac>
<mo>+</mo>
<msub>
<mi>t</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
<mrow>
<msub>
<mi>t</mi>
<mn>2</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>y</mi>
<mo>&times;</mo>
<mrow>
<mo>(</mo>
<msub>
<mi>t</mi>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>t</mi>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mn>255</mn>
</mfrac>
<mo>+</mo>
<msub>
<mi>t</mi>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
Wherein, t1,maxRepresent column vector t1The element of middle maximum, t1,minRepresent column vector t1The element of middle minimum, t2,maxRepresent row
Vectorial t2The element of middle maximum, t2,minRepresent column vector t2The element of middle minimum, p1To load vector paThe of=(1 ..., PC)
One row, p2To load vector paThe secondary series of=(1 ..., PC)；
Secondly, gray level image is calculated according to the following formula：
Lxy=[R, G, B]xy[0.299,0.587,0.114]T
Finally, flame brightness value is calculated according to the following formula：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>F</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</munder>
<msub>
<mi>TT</mi>
<mrow>
<mi>x</mi>
<mi>y</mi>
</mrow>
</msub>
<msub>
<mi>L</mi>
<mrow>
<mi>x</mi>
<mi>y</mi>
</mrow>
</msub>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mo>&ForAll;</mo>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>,</mo>
<msub>
<mi>M</mi>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>255</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein, TTx,yFor the number of similary pixel that coordinate in score block diagram is (x, y).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711293082.3A CN108052950B (en) | 2017-12-08 | 2017-12-08 | MIA-based fused magnesia furnace dynamic flame segmentation and feature extraction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711293082.3A CN108052950B (en) | 2017-12-08 | 2017-12-08 | MIA-based fused magnesia furnace dynamic flame segmentation and feature extraction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108052950A true CN108052950A (en) | 2018-05-18 |
| CN108052950B CN108052950B (en) | 2021-06-11 |
Family
ID=62122516
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711293082.3A Active CN108052950B (en) | 2017-12-08 | 2017-12-08 | MIA-based fused magnesia furnace dynamic flame segmentation and feature extraction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108052950B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109886231A (en) * | 2019-02-28 | 2019-06-14 | 重庆科技学院 | A kind of garbage burning factory Combustion Flame Recognition Using method |
| CN110057820A (en) * | 2019-04-15 | 2019-07-26 | 中南大学 | Method, system and the storage medium of on-line checking hydrogen chloride synthetic furnace chlorine hydrogen proportion |
| CN110660096A (en) * | 2019-10-08 | 2020-01-07 | 珠海格力电器股份有限公司 | Curve consistency detection method and storage medium |
| CN111598905A (en) * | 2020-05-13 | 2020-08-28 | 云垦智能科技(上海)有限公司 | Method for identifying type of blast furnace flame by using image segmentation technology |
| CN112669369A (en) * | 2021-01-20 | 2021-04-16 | 中国科学院广州能源研究所 | Quantitative determination method for degree of yellow flame of hydrocarbon flame |
| CN115880490A (en) * | 2022-11-21 | 2023-03-31 | 广东石油化工学院 | Flame segmentation method based on isolated forest |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1668920A (en) * | 2002-07-11 | 2005-09-14 | 麦克马斯特大学 | Method for online measurement of molten phases |
| CN102023160A (en) * | 2009-09-15 | 2011-04-20 | 财团法人工业技术研究院 | Measure method of combustible quality based on image |
| CN102598025A (en) * | 2009-08-25 | 2012-07-18 | 福瑞托-雷北美有限公司 | Method for real time detection of defects in a food product |
| CN103020496A (en) * | 2012-11-05 | 2013-04-03 | 王少夫 | Digital watermark encryption realization method |
| KR101309407B1 (en) * | 2012-11-20 | 2013-09-17 | 신현기 | The thermal and image of block image-based composite camera fire detector, fire detection system and method |
| CN103971114A (en) * | 2014-04-23 | 2014-08-06 | 天津航天中为数据***科技有限公司 | Forest fire detection method based on aerial remote sensing |
-
2017
- 2017-12-08 CN CN201711293082.3A patent/CN108052950B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1668920A (en) * | 2002-07-11 | 2005-09-14 | 麦克马斯特大学 | Method for online measurement of molten phases |
| CN102598025A (en) * | 2009-08-25 | 2012-07-18 | 福瑞托-雷北美有限公司 | Method for real time detection of defects in a food product |
| CN102023160A (en) * | 2009-09-15 | 2011-04-20 | 财团法人工业技术研究院 | Measure method of combustible quality based on image |
| CN103020496A (en) * | 2012-11-05 | 2013-04-03 | 王少夫 | Digital watermark encryption realization method |
| KR101309407B1 (en) * | 2012-11-20 | 2013-09-17 | 신현기 | The thermal and image of block image-based composite camera fire detector, fire detection system and method |
| CN103971114A (en) * | 2014-04-23 | 2014-08-06 | 天津航天中为数据***科技有限公司 | Forest fire detection method based on aerial remote sensing |
Non-Patent Citations (3)
| Title |
|---|
| G. SZATVANYI AND C. DUCHESNE，AT AL: "Multivariate Image Analysis of Flames for Product Quality and Combustion", 《AMERICAN CHEMICAL SOCIETY》 * |
| 李帷韬: "水泥回转窑烧成状态识别与熟料质量指标软测量的研究", 《中国博士学位论文全文数据库 工程科技Ⅰ辑》 * |
| 郭小玉: "炉口火焰序列图像颜色特征提取方法", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109886231A (en) * | 2019-02-28 | 2019-06-14 | 重庆科技学院 | A kind of garbage burning factory Combustion Flame Recognition Using method |
| CN110057820A (en) * | 2019-04-15 | 2019-07-26 | 中南大学 | Method, system and the storage medium of on-line checking hydrogen chloride synthetic furnace chlorine hydrogen proportion |
| CN110057820B (en) * | 2019-04-15 | 2021-04-30 | 中南大学 | Method, system and storage medium for on-line detection of chlorine-hydrogen ratio of hydrogen chloride synthesis furnace |
| CN110660096A (en) * | 2019-10-08 | 2020-01-07 | 珠海格力电器股份有限公司 | Curve consistency detection method and storage medium |
| CN111598905A (en) * | 2020-05-13 | 2020-08-28 | 云垦智能科技(上海)有限公司 | Method for identifying type of blast furnace flame by using image segmentation technology |
| CN112669369A (en) * | 2021-01-20 | 2021-04-16 | 中国科学院广州能源研究所 | Quantitative determination method for degree of yellow flame of hydrocarbon flame |
| CN115880490A (en) * | 2022-11-21 | 2023-03-31 | 广东石油化工学院 | Flame segmentation method based on isolated forest |
| CN115880490B (en) * | 2022-11-21 | 2023-10-27 | 广东石油化工学院 | Flame segmentation method based on isolated forest |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108052950B (en) | 2021-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108052950A (en) | A kind of segmentation of electric melting magnesium furnace dynamic flame and feature extracting method based on MIA | |
| JP5647878B2 (en) | Steel pipe internal corrosion analysis apparatus and steel pipe internal corrosion analysis method | |
| JP5721673B2 (en) | Paint color database creation method, creation system, creation program, and recording medium | |
| CN111047655B (en) | High-definition camera cloth defect detection method based on convolutional neural network | |
| US9305208B2 (en) | System and method for recognizing offensive images | |
| CN109903331A (en) | A kind of convolutional neural networks object detection method based on RGB-D camera | |
| CN110992238A (en) | Digital image tampering blind detection method based on dual-channel network | |
| JP2010097430A (en) | Smoke detection device and smoke detection method | |
| CN105741328A (en) | Shot image quality evaluation method based on visual perception | |
| CN104036493B (en) | No-reference image quality evaluation method based on multifractal spectrum | |
| CN107146220B (en) | A kind of universal non-reference picture quality appraisement method | |
| CN109657612A (en) | A kind of quality-ordered system and its application method based on facial image feature | |
| CN110400293A (en) | A kind of non-reference picture quality appraisement method based on depth forest classified | |
| KR20210141060A (en) | Machine learning based image anomaly detection system | |
| CN111666852A (en) | Micro-expression double-flow network identification method based on convolutional neural network | |
| CN114359733A (en) | Vision-based smoke fire detection method and system | |
| CN111833347A (en) | Transmission line damper defect detection method and related device | |
| CN111914938A (en) | Image attribute classification and identification method based on full convolution two-branch network | |
| CN107330441A (en) | Flame image foreground extraction algorithm | |
| Jakhetiya et al. | Distortion specific contrast based no-reference quality assessment of DIBR-synthesized views | |
| CN103020587B (en) | Based on video image analysis flame regarding figure analysis method | |
| Nugroho et al. | Negative content filtering for video application | |
| CN102930289A (en) | Method for generating mosaic picture | |
| Yang et al. | Blind image quality assessment on authentically distorted images with perceptual features | |
| CN111083468B (en) | Short video quality evaluation method and system based on image gradient |
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 |