CN103632373B - A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms - Google Patents
A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms Download PDFInfo
- Publication number
- CN103632373B CN103632373B CN201310658750.3A CN201310658750A CN103632373B CN 103632373 B CN103632373 B CN 103632373B CN 201310658750 A CN201310658750 A CN 201310658750A CN 103632373 B CN103632373 B CN 103632373B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- flco
- image
- frame
- 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.)
- Expired - Fee Related
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 49
- 239000002245 particle Substances 0.000 claims abstract description 25
- 230000002708 enhancing effect Effects 0.000 claims abstract description 16
- 238000000605 extraction Methods 0.000 claims abstract description 16
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 8
- 238000012805 post-processing Methods 0.000 claims abstract description 5
- 230000000694 effects Effects 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000005286 illumination Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 claims description 2
- 238000005189 flocculation Methods 0.000 claims description 2
- 230000016615 flocculation Effects 0.000 claims description 2
- 230000001788 irregular Effects 0.000 claims description 2
- 238000012804 iterative process Methods 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 5
- 239000000284 extract Substances 0.000 abstract description 4
- 238000004458 analytical method Methods 0.000 abstract description 3
- 238000002474 experimental method Methods 0.000 description 6
- 238000003672 processing method Methods 0.000 description 5
- 230000006872 improvement Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000005345 coagulation Methods 0.000 description 2
- 230000015271 coagulation Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 239000003651 drinking water Substances 0.000 description 1
- 235000020188 drinking water Nutrition 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 239000011121 hardwood Substances 0.000 description 1
- 238000003711 image thresholding Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000001954 sterilising effect Effects 0.000 description 1
- 238000004659 sterilization and disinfection Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms.The present invention is easily influenceed according to currently used flco tracking by noise, light and flco movement velocity etc., it is difficult to extract the feature of complete flco moving target, proposes that one kind is based on three-frame difference high-order statistic for this(HOS)Strengthen Da-Jin algorithm with reference to particle group optimizing(OTSU)Flco object detection method.This method carries out calculus of differences to continuous three two field picture first, then Fourth-order moment is calculated pixel-by-pixel and compared with threshold value, wherein optimal threshold is obtained using particle group optimizing enhancing Da-Jin algorithm, image binaryzation is carried out using optimal threshold and carries out post processing of image, more visible flco target is finally obtained, is laid the foundation for follow-up flco is analysis automated.The inventive method has quickly and accurately feature, can effectively extract flco target, effective extraction of flco target suitable for water process.
Description
Technical field
The present invention relates to a kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms, category water process wadding
Body detecting method technical field.
Background technology
With the continuous improvement of people's living standards, requirement of the people to drinking water quality also more and more higher, both at home and abroad often
Rule water treatment technology generally comprises several stages such as coagulation, precipitation, filtering, sterilization.Wadding can be produced during water treatment
Body condenses phenomenon, and flco number, size, sinking speed etc. be the important parameter for judging coagulation effect.Moving target is examined
Survey is the essential step in floc image quantitative analysis, and influences analysis automated flco, job stability, result accuracy
Key.Uneven illumination is even, flco movement velocity is slow, wadding due to there may be in imaging process for industrial camera acquired image
Situations such as body changes in distribution so that the shade of flco differs, lack of homogeneity, while noise be present, and flco is also present
The different phenomenon of surface reflection.Therefore the detection of flco target and background is realized exactly, is further analysis floc characteristic
Basis.
Realize that flco mesh object detection method has much at present, the conventional calculus of finite differences that has powerful connections, frame differential method and light stream
Method.
Frame differential method is 2 or 3 adjacent interframe in the video sequence, is carried using difference pixel-by-pixel and thresholding
The moving region in image is taken, it can detect the part to be changed in video image faster.Frame difference method method is simple, right
The adaptability of environment is relatively good, and stability is high, in the ideal case, when video camera is static, if subtracting each other gray scale between rear continuous hardwood
Difference is zero, then it is assumed that the point belongs to static background, on the contrary then belong to moving object region.But often exist in actual conditions a lot
Noise jamming, only whether it is zero situation of change that not can determine that the point according to difference.Three-frame difference is using three two field pictures difference phase
Subtract, effect is better than two frame differences.Simply though error image thresholding method can be gone out move mesh with substantial separation in frame difference method
Mark and background, but the establishing method of threshold value is often more difficult, and be difficult the influence for filtering out noise.So need further
Improve and moving target and background are efficiently separated to the choosing method of threshold value to the processing method of noise and improvement.
For method for processing noise, currently there are many processing methods, such as smoothing denoising, wherein using high-order statistic
(higher order statistics, HOS), refer to the statistic of 3 ranks or 3 rank above exponent numbers, including Higher Order Cumulants, height
Rank square and higher-order spectrum, Higher Order Cumulants can completely inhibit the influence of Gaussian noise and some other characteristics in theory,
Therefore, generally more it is used as processing noise instrument by the use of Higher Order Cumulants and higher-order spectrum.
For Research on threshold selection, mainly there are Two-peak method, iterative method, maximum variance between clusters (OTSU, i.e. Da-Jin algorithm).It is double
Although peak method is simple, adaptability is poor;Iterative method operand is huge, is not suitable for real-time system;Comparatively Da-Jin algorithm
Effect it is best.It is two classes that Da-Jin algorithm, which is first divided to image pixel threshold value T, then calculates the inter-class variance of two class grey scale pixel values
And variance within clusters, carry out threshold value T during using ratio between two to be maximum.
But find in simulated experiment, when handling background and foreground target gray scale is more or less the same, occurred with Da-Jin algorithm
Substantial amounts of black region, even it can lose whole target when serious.There is scholar to propose to introduce a kind of big Tianjin of the enhancing of gray scale stretching
Method, i.e., gray scale stretching is carried out to image using nonlinear transformation, it is then determined that threshold value, but which increases operation time.
Particle swarm optimization algorithm is a kind of probabilistic search method found during being simulated to group behavior.Calculate
Method passes through the optimal solution of iterative search particle fitness function since primary group.In each iteration, each particle
Optimal solution gbest that the optimal solution pbest and whole population found according to itself is found adjusts the speed of motion and side
To with the position of more new particle.
Particle group optimizing enhancing Da-Jin algorithm (OTSU) is exactly that enhancing Da-Jin algorithm is optimized using particle swarm optimization algorithm,
It effectively solves the matter of time of enhancing Da-Jin algorithm, and a RANDOM SOLUTION is given in position and speed of the algorithm first to particle, by
Then gray level image scope is between 0-255, so to give particle position within this range, all random values are by rand
() function produces.Then calculate fitness function value and be compared, iterate to calculate, finally try to achieve optimal solution, that is, most preferably
Threshold value.
The content of the invention
The present invention seeks to ask existing for the deficiency and influence of noise and present threshold value choosing method for Three image difference
Topic, the present invention propose one kind and divide high-order statistic (HOS) and particle group optimizing to strengthen Da-Jin algorithm (OTSU) based on three-frame difference
Flco object detection method.
Realize the technical scheme is that, the present invention based on motion flco it is irregular, it is affected by environment big the reason for, make
With based on three-frame difference, high-order statistic and particle group optimizing enhancing Da-Jin algorithm extraction floc characteristic, comprise the following steps:
(1) respectively to present frame It, former frame It-1, a later frame It+1Pre-processed, including gray-level correction, image enhaucament
Deng processing;
(2)It-1, It, It+1Difference is done respectively, obtains difference image Dt+1, Dt;
(3) difference image is carried out mutually obtaining difference image D with computingi;
(4) optimal threshold T is determined using particle group optimizing enhancing Da-Jin algorithm;
(5) Fourth-order moment and compared with optimal threshold T is calculated pixel-by-pixel, more than T values, it is believed that it is moving target, pixel value
1 is set to, is otherwise 0;
(6) to the imagery exploitation Image Post-processing Techniques in step (5), incomplete border is further improved, so as to
To complete boundary image, so as to realize the smooth extraction of flco target.
The method for determining optimal threshold, a RANDOM SOLUTION is given in the position to particle and speed first, due to being gray scale
Image range is between 0-255, so to give particle position within this range, all random values are produced by rand () function
It is raw.Then calculate fitness function value and be compared, iterate to calculate, finally try to achieve optimal solution, that is, optimal threshold.
Described image post-processing technology, i.e., a "ON" computing (Open is first to image of less structural element
Operation isolated noise) is eliminated;Make once " closing " computing (Close to image of larger structural element again
Operation target internal cavity) is filled up;Incomplete border is further improved, so as to obtain complete boundary image, so as to
Realize the smooth extraction of flco target.
The present invention is based on flco motion characteristics, it is proposed that excellent based on three-frame difference high-order statistic (HOS) and population
Change the method that enhancing Da-Jin algorithm carries out flco detection and extraction.Compared with existing scheme, the improved method of the present invention can compare
The edge of motion flco is accurately extracted, the flco target area of extraction is also than more complete, mesh especially larger to area
Mark the deficiency that extraction effect is good, solves traditional two frame difference methods to a certain extent, and noise jamming, side in detection process
Edge detection is imperfect, is also easy to produce the shortcomings that target " cavity ", can accurately realize the extract real-time of flco target area.
The invention has the advantages that the present invention avoids two frame differences using three-frame difference method extraction flco target
The problems such as Objective extraction that method is likely to occur is imperfect;But the On The Choice of optimal threshold in three-frame difference method simultaneously be present
With influence of noise problem, the present invention chooses optimal threshold using population enhancing Da-Jin algorithm, efficiently solved using iterative method etc.
The threshold value that method is likely to occur chooses unreasonable, the problems such as efficiency is low, and uses HOS technologies, avoids the interference of noise.Cause
This, improved method proposed by the present invention has quickly, accurately, extracts the advantages that flco target is complete, and noiseproof feature is good, the invention
There is certain practical value, can effectively extract flco target.
The present invention moves the Objective extraction of flco suitable for water process.
Brief description of the drawings
Fig. 1 is the preferable experimental facilities schematic diagram of the present invention;
Fig. 2 is the flow chart of the present invention;
Fig. 3 is the contrast effect figure that method emulates in the present invention:
Fig. 3 (a) is original image;
Fig. 3 (b) is original image;
Fig. 3 (c) is original image;
Fig. 3 (d) is two three-frame differences to original image Fig. 3 (a);
Fig. 3 (e) is two three-frame differences to original image Fig. 3 (b);
Fig. 3 (f) is two three-frame differences to original image Fig. 3 (c);
Fig. 3 (g) is three three-frame differences+iteration to original image Fig. 3 (a);
Fig. 3 (h) is three three-frame differences+iteration to original image Fig. 3 (b);
Fig. 3 (i) is three three-frame differences+iteration to original image Fig. 3 (c);
Fig. 3 (j) is three three-frame differences+Da-Jin algorithm to original image Fig. 3 (a);
Fig. 3 (k) is three three-frame differences+Da-Jin algorithm to original image Fig. 3 (b);
Fig. 3 (l) is three three-frame differences+Da-Jin algorithm to original image Fig. 3 (c);
Fig. 3 (m) is the processing method of the present invention to original image Fig. 3 (a);
Fig. 3 (n) is the processing method of the present invention to original image Fig. 3 (b);
Fig. 3 (o) is the processing method of the present invention to original image Fig. 3 (c).
Embodiment
According to the step of the inventive method, the specific embodiment of the invention is as follows:
The first step:According to shown in Fig. 1, experiment equipment is put.The install sensor in the water of flocculation basin end;Make current water
Put down, slowly flow through sampling window, by current (flco) image of industry camera continuous acquisition sampling window, basis in experiment
Actual demand sets sampling time interval.In this experiment time interval Tsampe for using for 1s (within the 1s times, observation window
The flco number of body has good representativeness) population be 100 or so, can be used as sample be used for handle image.
Second step:Three-frame difference and noise processed.Respectively to the former frame f of the image of collectiont-1(x, y), present frame ft
(x, y), a later frame ft+1(x, y) carries out smoothing denoising, then calculates ft-1(x, y) and ft(x, y) difference image Dt, ft+1(x,y)
With ft(x, y) difference image Dt+1, then by Dt, Dt+1Carry out with obtaining difference image Q after computingdiff(x, y), finally choose and close
Suitable threshold value, for Three image difference, chooses rational threshold value and the effect of Objective extraction is played an important role, according to formula (6)
Thresholding is carried out to difference result and obtains binary result.;
(A) under conditions of assuming that extraneous illumination condition does not change or changed less, the expression of adjacent image sequence is as follows:
ft(x, y)=Mt(x,y)+Bt(x,y)+nt(x,y) (1)
(B) two formulas obtain adjacent two field pictures difference image after subtracting each other is:
Present frame and the difference image of a later frame are calculated by above-mentioned formula, and calculates the difference of present frame and former frame according to this
Partial image.
(C) Fourth-order moment of each pixel, (x, y) point Fourth-order moment average value are detectedFor,
Wherein, η (x, y) takes 3 × 3 moving window centered on (x, y),For frame-to-frame differences ash in moving window
The average value of angle value:
(D) binaryzation
Appropriate threshold is chosen, for Three image difference, rational threshold value is chosen and important work is played to the effect of Objective extraction
With below with optimization enhancing Da-Jin algorithm threshold value.
3rd step:The selection of threshold value in second step.It is that particle group optimizing enhancing Da-Jin algorithm is true using the Da-Jin algorithm after improvement
Determine optimal threshold.Enhancing Da-Jin algorithm due to introduce gray scale stretching improve Da-Jin algorithm effect while also increase computing when
Between, when the threshold value applied to real-time online is chosen, it is unfavorable for real-time operation, time-based requirement, the present embodiment is big to strengthening
Tianjin method is improved, i.e., optimizes enhancing Da-Jin algorithm using particle swarm optimization algorithm, and the position to particle and speed first gives one
RANDOM SOLUTION, due to be gray level image scope between 0-255, so to give particle position within this range, all is random
Value is produced by rand () function.Then calculate fitness function value and be compared, iterate to calculate, finally try to achieve optimal solution,
It is exactly optimal threshold.
Weight w and c in calculating process1,c2Selection etc. parameter is very crucial, and wherein weight w selection influence method is searched
Suo Nengli.Under normal circumstances, it is desirable to have stronger ability of searching optimum initial when, stage lays particular stress on local search ability
Enhancing;W values are from 0.9 linear decrease to 0.4 typically in search procedure, in order that particle can preferably find optimal value,
Typically take c1=c2=2.
4th step:The optimal threshold in formula (6) and second step in the first step carries out binaryzation to difference result,
Determine binary image.
5th step:To the imagery exploitation image procossing post-processing technology in (5), such as:1) first with less structural element B1
A "ON" computing (Open Operation) is done to image and eliminates isolated noise;2) again with larger structural element B2 to image
Do once " closing " computing (Close Operation) and fill up target internal cavity;Incomplete border is further improved, so as to
To complete boundary image, so as to realize the smooth extraction of flco target.
6th step:Software emulation realizes that Fig. 3 is simulation result.With 3 kinds of common methods and the inventive method, distinguish
Into the design and experiment of motion detection block, whole program is based on Windows operating system, employs mathwork companies
Matlab7.0 or Microsoft Visual C++6.0 are programming development platform.
Cameras configuration used in the present embodiment is CCD Techno-Industrial video cameras, has used the video sequence of 10 groups of shootings
Row image is tested, and each sequence contains 30 frames or so, tiff forms that image size is 312 × 288,256 × 190
Gif format-patterns etc., representational continuous three two field picture is taken respectively, with common 3 kinds of methods (two three-frame difference methods, three
Three-frame difference method+alternative manner, three three-frame differences+Da-Jin algorithm) and method of the invention, it has been respectively completed motion detection block
Design and experiment, whole program is based on Windows operating system, employ mathwork companies matlab7.0 and
Microsoft Visual C++6.0 are programming development platform.
Claims (1)
- A kind of 1. flco detection method of three-frame difference high-order statistic combination OTSU algorithms, it is characterised in that methods described base In motion flco it is irregular, it is affected by environment big the reason for, using based on three-frame difference, high-order statistic and particle group optimizing increase Strong Da-Jin algorithm OTSU extraction floc characteristics, comprise the following steps:The first step:The install sensor in the water of flocculation basin end;Make current level, slowly flow through sampling window, pass through industry The current floc image of camera continuous acquisition sampling window;Second step:Respectively to the former frame f of the image of collectiont-1(x, y), present frame ft(x, y), a later frame ft+1(x, y) is put down Sliding denoising, then calculates ft-1(x, y) and ftThe difference image of (x, y), ft+1(x, y) and ftThe difference image of (x, y);(A) under conditions of assuming that extraneous illumination condition does not change or changed less, the expression of adjacent image sequence is as follows:ft(x, y)=Mt(x,y)+Bt(x,y)+nt(x,y) (1)<mrow> <msub> <mi>f</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>=</mo> <msub> <mi>M</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>+</mo> <mover> <mi>x</mi> <mo>&OverBar;</mo> </mover> <mo>,</mo> <mi>y</mi> <mo>+</mo> <mover> <mi>y</mi> <mo>&OverBar;</mo> </mover> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>B</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>n</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow>(B) two formulas obtain adjacent two field pictures difference image after subtracting each other is:<mrow> <mtable> <mtr> <mtd> <mrow> <mi>d</mi> <mrow> <mo>(</mo> <mi>s</mi> <mo>,</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <msub> <mi>f</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>f</mi> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>=</mo> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>M</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>+</mo> <mover> <mi>x</mi> <mo>&OverBar;</mo> </mover> <mo>,</mo> <mi>y</mi> <mo>+</mo> <mover> <mi>y</mi> <mo>&OverBar;</mo> </mover> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>M</mi> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>B</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>B</mi> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>n</mi> <mrow> <mi>t</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>-</mo> <msub> <mi>n</mi> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> </mtable> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </mrow>Present frame and the difference image of a later frame are calculated by above-mentioned formula, and calculates present frame and the difference diagram of former frame according to this Picture;Then above-mentioned two difference images being calculated are carried out with obtaining difference image Q after computingdiff(x,y);(C) Fourth-order moment of each pixel, (x, y) point Fourth-order moment average value are detectedFor,Wherein, η (x, y) takes 3 × 3 moving window centered on (x, y),For frame-to-frame differences gray value in moving window Average value:(D) binaryzationAppropriate threshold is chosen, for Three image difference, rational threshold value is chosen and the effect of Objective extraction is played an important role, under Face optimization enhancing Da-Jin algorithm threshold value T;3rd step:The selection of threshold value in second step:Enhancing Da-Jin algorithm, the first position to particle are optimized using particle swarm optimization algorithm Put with speed to a RANDOM SOLUTION, due to be gray level image scope between 0-255, so to give particle position within this range Put, all random values are produced by rand () function, are then calculated fitness function value and are compared, iterate to calculate, finally Try to achieve optimal solution, that is, optimal threshold T;In the iterative process, weight w values from 0.9 linear decrease to 0.4, in order that Particle can preferably find optimal value, take c1=c2=2;4th step:The optimal threshold T in formula (6) and the 3rd step in second step is to difference result Qdiff(x, y) carries out two Value, binary image is determined, more than or equal to T values, it is believed that it is moving target, pixel value set 1, is otherwise 0;5th step:Using image procossing post-processing technology, incomplete border is further improved, so as to obtain complete boundary graph Picture, so as to realize the smooth extraction of flco target.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310658750.3A CN103632373B (en) | 2013-12-09 | 2013-12-09 | A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310658750.3A CN103632373B (en) | 2013-12-09 | 2013-12-09 | A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103632373A CN103632373A (en) | 2014-03-12 |
CN103632373B true CN103632373B (en) | 2017-12-22 |
Family
ID=50213390
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310658750.3A Expired - Fee Related CN103632373B (en) | 2013-12-09 | 2013-12-09 | A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103632373B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106997598A (en) * | 2017-01-06 | 2017-08-01 | 陕西科技大学 | The moving target detecting method merged based on RPCA with three-frame difference |
CN109828236A (en) * | 2019-02-14 | 2019-05-31 | 中南大学 | A kind of microseism/acoustic emission source locating method in labyrinth containing dead zone |
CN111833269B (en) * | 2020-07-13 | 2024-02-02 | 字节跳动有限公司 | Video noise reduction method, device, electronic equipment and computer readable medium |
CN112418105B (en) * | 2020-11-25 | 2022-09-27 | 湖北工业大学 | High maneuvering satellite time sequence remote sensing image moving ship target detection method based on difference method |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102184550A (en) * | 2011-05-04 | 2011-09-14 | 华中科技大学 | Mobile platform ground movement object detection method |
-
2013
- 2013-12-09 CN CN201310658750.3A patent/CN103632373B/en not_active Expired - Fee Related
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102184550A (en) * | 2011-05-04 | 2011-09-14 | 华中科技大学 | Mobile platform ground movement object detection method |
Non-Patent Citations (1)
Title |
---|
嵌入式视频跟踪算法的研究;张之稳;《中国优秀硕士学位论文全文数据库》;20061215;正文第4章 * |
Also Published As
Publication number | Publication date |
---|---|
CN103632373A (en) | 2014-03-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107204006B (en) | Static target detection method based on double background difference | |
Rong et al. | An improved CANNY edge detection algorithm | |
CN103268480B (en) | A kind of Visual Tracking System and method | |
CN108230264B (en) | Single image defogging method based on ResNet neural network | |
CN102568005B (en) | Moving object detection method based on Gaussian mixture model | |
Huang et al. | An advanced single-image visibility restoration algorithm for real-world hazy scenes | |
CN104392468B (en) | Based on the moving target detecting method for improving visual background extraction | |
CN102307274B (en) | Motion detection method based on edge detection and frame difference | |
CN106407927B (en) | The significance visual method suitable for underwater target detection based on polarization imaging | |
CN104268872B (en) | Consistency-based edge detection method | |
CN103955949B (en) | Moving target detecting method based on Mean-shift algorithm | |
CN102982537B (en) | A kind of method and system detecting scene change | |
CN103632373B (en) | A kind of flco detection method of three-frame difference high-order statistic combination OTSU algorithms | |
CN106157323A (en) | The insulator division and extracting method that a kind of dynamic division threshold value and block search combine | |
Feng et al. | A separating method of adjacent apples based on machine vision and chain code information | |
CN106485702A (en) | Image blurring detection method based on natural image characteristic statisticses | |
CN105405138A (en) | Water surface target tracking method based on saliency detection | |
CN105787912A (en) | Classification-based step type edge sub pixel localization method | |
CN111860143A (en) | Real-time flame detection method for inspection robot | |
CN107871315B (en) | Video image motion detection method and device | |
CN113610024A (en) | Multi-strategy deep learning remote sensing image small target detection method | |
CN101877135B (en) | Moving target detecting method based on background reconstruction | |
CN109241932A (en) | A kind of thermal infrared human motion recognition method based on movement variogram phase property | |
CN105844671B (en) | A kind of fast background relief method under the conditions of change illumination | |
CN105741317B (en) | Infrared motion target detection method based on time-space domain significance analysis and rarefaction representation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171222 |
|
CF01 | Termination of patent right due to non-payment of annual fee |