CN107274391A - A kind of underwater picture Objective extraction system - Google Patents
A kind of underwater picture Objective extraction system Download PDFInfo
- Publication number
- CN107274391A CN107274391A CN201710424076.0A CN201710424076A CN107274391A CN 107274391 A CN107274391 A CN 107274391A CN 201710424076 A CN201710424076 A CN 201710424076A CN 107274391 A CN107274391 A CN 107274391A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- image
- local area
- pixel
- 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- Withdrawn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
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/181—Segmentation; Edge detection involving edge growing; involving edge linking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- 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
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
Image object extraction system in a kind of water, including main control module, underwater robot module, image processing module, information transmission modular and display module, the main control module is used for the running for controlling underwater robot module and the collection of original image under water, described image processing module is used to handle the original image, so as to recognize the target in image, described information transport module is shown for the image information after processing to be transferred into display module.Beneficial effects of the present invention are：The system carries out effective collection of underwater picture by underwater robot module, and subarea processing is carried out to the image collected, the interference such as the effective noise eliminated in original image, and then by carrying out edge treated and centre of form extraction to filtered image, realize the quick identification of submarine target.
Description
Technical field
The invention is related to object detection field, and in particular to a kind of submarine target extraction system.
Background technology
In recent years, continuing to develop with international situation, ocean is increasingly becoming the new strategic center of gravity in countries in the world, ocean phase
The research in pass field is important for all having in terms of marine resources detection and exploitation, marine environmental monitoring and ocean Military Application
Meaning.Therefore, the present invention provides a kind of underwater picture Objective extraction system, realizes the effective detection of submarine target, improves
The accuracy of underwater target detection.
The content of the invention
In view of the above-mentioned problems, a kind of the present invention is intended to provide Objective extraction system in water.
The purpose of the invention is achieved through the following technical solutions：
Image object extraction system in a kind of water, including main control module, underwater robot module, image processing module, letter
Transport module and display module are ceased, the main control module is used for the running for controlling underwater robot and control machine people is carried out under water
The collection of original image, described image processing module is used to handle the original image, so as to recognize the mesh in image
Mark, described information transport module is shown for the image information after processing to be transferred into display module.
The beneficial effect of the invention：The system carries out effective collection of underwater picture by underwater robot module,
And subarea processing is carried out to the image collected, effectively eliminate noise in original image etc. and disturb, and then pass through
Edge treated is carried out to filtered image and the centre of form is extracted, the quick identification of submarine target is realized.
Brief description of the drawings
Innovation and creation are described further using accompanying drawing, but the embodiment in accompanying drawing does not constitute and the invention is appointed
What is limited, on the premise of not paying creative work, can also be according to the following drawings for one of ordinary skill in the art
Obtain other accompanying drawings.
Fig. 1 is schematic structural view of the invention；
Fig. 2 is underwater robot modular structure schematic diagram of the present invention
Fig. 3 is image processing module structural representation of the present invention.
Reference：
Main control module 1；Underwater robot module 2；Image processing module 3；Information transmission modular 3；Display module 4；Control
Device unit 21；Underwater camera unit 22；Underwater lighting unit 23；Gaussian filtering unit 31；Edge treated unit 32；The centre of form is extracted
Unit 33.
Embodiment
The invention will be further described with the following Examples.
Referring to Fig. 1, Fig. 2 and Fig. 3, image object extraction system in a kind of water of the present embodiment, including main control module 1, water
Lower robot module 2, image processing module 3, information transmission modular 4 and display module 5, the main control module 1 are used to control water
The running of lower robot module 2 and the under water collection of original image, described image processing module 3 are used to enter the original image
Row processing, so as to recognize the target in image, described information transport module 4 is used to the image information after processing being transferred to display
Module 5 is shown.
Preferably, the underwater robot module 2 includes controller unit 21, underwater camera unit 22 and underwater lighting list
Member 23, the controller unit 21 receives the instruction from main control module 1 by information transmission modular 4, so as to be adjusted according to instruction
The running status of whole robot, the underwater camera unit 22 is used to gather original image under water, the underwater lighting unit
23 are used to be illuminated when image unit 22 is gathered under water.
Preferably, described information transport module 4 is communicated using water-proof cable with main control module 1.
This preferred embodiment carries out effective collection of underwater picture, and the figure to collecting by underwater robot module
As carrying out subarea processing, effectively eliminate noise in original image etc. and disturb, and then by entering to filtered image
Row edge treated and the centre of form are extracted, and realize the quick identification of submarine target.
Preferably, described image processing module 3 includes gaussian filtering unit 31, edge treated unit 32 and centre of form extraction list
Member 33, the gaussian filtering unit 31 is used to handle the original image collected, removes noise in image etc. and does
Disturb, and realize being accurately positioned for image border, the edge treated unit 32 is used to handle the edge of filtered image,
Remaining pseudo-edge in filtered image is removed, the profile that the centre of form extraction unit 33 is used for the image according to obtained by detection is carried
The centre of form of target is taken, so as to realize the quick identification of target.
Preferably, the gaussian filtering unit 31 is used to carry out at gaussian filtering the original image under water collected
Reason, it uses a kind of improved filter scale algorithm to carry out the determination of gaussian filtering scale parameter, specifically included：
A. subregion filtering is carried out to the original image, calculates the combinatorial complexity factor of the regional area of image, its
Computational methods are as follows：
In formula, ExFor image local area DxPixel average, ρxFor image local area DxPixel standard
Variance, (m, n) is image local area DxCentral point, s be image local area DxRanks value, XijFor image local area
DxThe pixel at position (i, j) place；
In formula, δxFor image local area DxInformation entropy, h is total gray level of image, ziFor image local area
DxMiddle gray level is i pixel sum, piFor image local area DxMiddle gray level is the probability of i pixel；
Define image local area DxPosition (mi,ni) place pixel be f (mi,ni), then image local area DxIn
Pixel f (mi,ni) θ directions average gradient value be ωx；
In formula, g is image local area DxPixel sum,For image local area DxPosition (mi,ni) place's picture
The abscissa value of vegetarian refreshments,For image local area DxPosition (mi,ni) place's pixel ordinate value；
Then image local area DxCombinatorial complexity factor fxCalculation formula it is as follows：
In formula, δxFor the information entropy of image local area, ρxFor image local area DxPixel standard variance,
ωxFor the average gradient value of image local area,WithRespectively image local area information entropy, standard variance and
Average gradient value is in combinatorial complexity factor fxIn shared ratio；
B. according to the combinatorial complexity factor of the image local area of above-mentioned gained calculate regional area filter scale because
Sub- px, then regional area DxFilter scale factor pxCalculation formula be：
pmax=log2[max(H,L)]
In formula, fxIt is image local area DxLocal Complexity, fmaxIt is the maximum Local Complexity in regional area,
fminIt is the minimum Local Complexity of image local area, pmaxIt is the filter scale factor of Local Minimum complexity, (H, L) is office
The resolution ratio of portion's least complex correspondence image regional area；
C. according to the filter scale factor p of each regional area of above-mentioned calculating gainedx, calculate the gaussian filtering ginseng of corresponding region
Number σxValue be：
This preferred embodiment has introduced the standard deviation of regional area, letter in the calculating process of image local area complexity
The combination of entropy and average gradient value is ceased as the count factor of complexity, adds the standard of the Local Complexity result of calculation of image
True property, carries out regional area filtering to the original image under water collected, is defined according to the Local Complexity of each regional area
The filter scale size in each region, realizes the filtering that the different zones of underwater picture are carried out with different scale.
Preferably, the edge treated unit 32 is used to handle filtered image, removes in filtered image
Remaining pseudo-edge, is specifically included：
A. the Grad of filtered image is calculated, so as to set up the histogram of gradients of filtered image, gradient Nogata is defined
The most gradient magnitude of pixel number is T in figurem, define the high threshold y of edge connectionm, then threshold value ymCalculation formula be：
In formula, TiFor the gradient magnitude of image, TmFor the most gradient magnitude of pixel number in histogram of gradients, N is image
The number of middle Grad；
B. remove and be more than above-mentioned gained high threshold y in histogram of gradientsmPixel, the gradient for setting up residual pixel point is straight
Fang Tu, it is K to define the most gradient magnitude of pixel number in the histogram of gradients of residual pixel pointm, define the low of edge connection
Threshold value yl, then Low threshold ylCalculation formula be：
In formula, KiFor the Grad in the histogram of gradients of residual pixel point, M is in the histogram of gradients of residual pixel point
The number of gradient magnitude, KmFor the most gradient magnitude of pixel number in the histogram of gradients of residual pixel point；
C. remove and be more than threshold value y in histogram of gradientslPixel, remaining pixel is the edge of image.
This preferred embodiment is according to the adaptive threshold value of the histogrammic statistical information of image gradient, with the fixed threshold of tradition
The method of value is compared, it is to avoid lose gray level value changes slow local edges, it is ensured that the accuracy of edge detection results.
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than to present invention guarantor
The limitation of scope is protected, although being explained with reference to preferred embodiment to the present invention, one of ordinary skill in the art should
Work as understanding, technical scheme can be modified or equivalent substitution, without departing from the reality of technical solution of the present invention
Matter and scope.
Claims (6)
1. image object extraction system in a kind of water, it is characterized in that, including main control module, underwater robot module, image procossing
Module, information transmission modular and display module, the main control module are used for the running for controlling underwater robot module and former under water
The collection of beginning image, described image processing module is used to handle the original image, so that the target in image is recognized,
Described information transport module is shown for the image information after processing to be transferred into display module.
2. image object extraction system in a kind of water according to claim 1, it is characterized in that, the underwater robot module
Including controller unit, underwater camera unit and underwater lighting unit, the controller unit is received by information transmission modular
Instruction from main control module, so as to adjust the running status of robot according to instruction, the underwater camera unit is used to gather
Original image under water, the underwater lighting unit is used to be illuminated when camera unit is gathered under water.
3. image object extraction system in a kind of water according to claim 2, it is characterized in that, described information transport module is adopted
Communicated with water-proof cable with main control module.
4. image object extraction system in a kind of water according to claim 3, it is characterized in that, described image processing module bag
Gaussian filtering unit, edge treated unit and centre of form extraction unit are included, the gaussian filtering unit is used for the original to collecting
Beginning image is handled, and is removed noise in image etc. and is disturbed, and realizes being accurately positioned for image border, the edge treated list
Member is used to handle the edge of filtered image, removes remaining pseudo-edge in filtered image, and the centre of form is extracted single
Member is used for the centre of form of the contours extract target of the image according to obtained by detection, so as to realize the quick identification of target.
5. image object extraction system in a kind of water according to claim 4, it is characterized in that, the gaussian filtering unit is used
Gaussian filtering process is carried out in the original image under water to collecting, it uses a kind of improved filter scale algorithm to carry out
The determination of gaussian filtering scale parameter, is specifically included：
A. subregion filtering is carried out to the original image, calculates the combinatorial complexity factor of the regional area of image, it is calculated
Method is as follows：
<mrow>
<msub>
<mi>E</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mi>m</mi>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mrow>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
<mrow>
<mi>m</mi>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
</msubsup>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mi>n</mi>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mrow>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
<mrow>
<mi>n</mi>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
</msubsup>
<msub>
<mi>X</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
</mrow>
<msup>
<mi>S</mi>
<mn>2</mn>
</msup>
</mfrac>
</mrow>
<mrow>
<msub>
<mi>&rho;</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mfrac>
<msqrt>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mi>m</mi>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mrow>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
<mrow>
<mi>m</mi>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
</msubsup>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mi>n</mi>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mrow>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
<mrow>
<mi>n</mi>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mi>s</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
</msubsup>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>E</mi>
<mi>x</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mi>S</mi>
</mfrac>
</mrow>
In formula, ExFor image local area DxPixel average, ρxFor image local area DxPixel standard variance,
(m, n) is image local area DxCentral point, s be image local area DxRanks value, XijFor image local area DxPosition
The pixel at (i, j) place；
<mrow>
<msub>
<mi>&delta;</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mo>-</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>h</mi>
</munderover>
<msub>
<mi>p</mi>
<mi>i</mi>
</msub>
<mo>*</mo>
<msub>
<mi>logp</mi>
<mi>i</mi>
</msub>
</mrow>
<mrow>
<msub>
<mi>p</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>Z</mi>
<mi>i</mi>
</msub>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>h</mi>
</msubsup>
<msub>
<mi>z</mi>
<mi>i</mi>
</msub>
</mrow>
</mfrac>
</mrow>
In formula, δxFor image local area DxInformation entropy, h is total gray level of image, ziFor image local area DxIn
Gray level is i pixel sum, piFor image local area DxMiddle gray level is the probability of i pixel number；
Define image local area DxPosition (mi,ni) place pixel be f (mi,ni), then image local area DxPosition (mi,
ni) pixel f (m in placei,ni) θ directions average gradient value be ωx；
<mrow>
<msub>
<mi>&omega;</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mrow>
<mo>(</mo>
<msub>
<mi>f</mi>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
</msub>
<mi>c</mi>
<mi>o</mi>
<mi>s</mi>
<mi>&theta;</mi>
<mo>+</mo>
<msub>
<mi>f</mi>
<msub>
<mi>n</mi>
<mi>i</mi>
</msub>
</msub>
<mi>s</mi>
<mi>i</mi>
<mi>n</mi>
<mi>&theta;</mi>
<mo>)</mo>
</mrow>
</mrow>
<mi>g</mi>
</mfrac>
</mrow>
1
In formula, g is image local area DxPixel sum,For image local area DxPosition (mi,ni) place's pixel
Abscissa value,For image local area DxPosition (mi,ni) place's pixel ordinate value；
Then image local area DxCombinatorial complexity factor fxCalculation formula it is as follows：
<mrow>
<msub>
<mi>f</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<msub>
<mi>&theta;</mi>
<mn>1</mn>
</msub>
<msub>
<mi>&delta;</mi>
<mi>x</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&theta;</mi>
<mn>2</mn>
</msub>
<msub>
<mi>&omega;</mi>
<mi>x</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&theta;</mi>
<mn>3</mn>
</msub>
<msup>
<mi>ln</mi>
<mrow>
<mo>(</mo>
<mrow>
<mn>1</mn>
<mo>+</mo>
<msub>
<mi>&rho;</mi>
<mi>x</mi>
</msub>
</mrow>
<mo>)</mo>
</mrow>
</msup>
</mrow>
In formula, δxFor the information entropy of image local area, ρxFor image local area DxPixel standard variance, ωxFor
The average gradient value of image local area, θ1、θ2And θ3Respectively image local area information entropy, standard variance and average ladder
Angle value is in combinatorial complexity factor fxIn shared ratio；
B. the filter scale factor p of regional area is calculated according to the combinatorial complexity factor of the image local area of above-mentioned gainedx,
Then regional area DxFilter scale factor pxCalculation formula be：
<mrow>
<msub>
<mi>p</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>p</mi>
<mi>max</mi>
</msub>
<mo>;</mo>
<msub>
<mi>f</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<msub>
<mi>f</mi>
<mi>min</mi>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>p</mi>
<mi>max</mi>
</msub>
<mrow>
<mo>(</mo>
<mrow>
<mn>1</mn>
<mo>-</mo>
<mfrac>
<msub>
<mi>f</mi>
<mi>x</mi>
</msub>
<mrow>
<msub>
<mi>f</mi>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>f</mi>
<mi>min</mi>
</msub>
</mrow>
</mfrac>
</mrow>
<mo>)</mo>
</mrow>
<mo>;</mo>
<msub>
<mi>f</mi>
<mi>min</mi>
</msub>
<mo><</mo>
<msub>
<mi>f</mi>
<mi>x</mi>
</msub>
<mo><</mo>
<msub>
<mi>f</mi>
<mi>max</mi>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>;</mo>
<msub>
<mi>f</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<msub>
<mi>f</mi>
<mi>max</mi>
</msub>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
pmax=log2[max(H,L)]
In formula, fxIt is image local area DxLocal Complexity, fmaxIt is the maximum Local Complexity in regional area, fminIt is
The minimum Local Complexity of image local area, pmaxIt is the filter scale factor of Local Minimum complexity, (H, L) is minimum office
The resolution ratio of portion's complexity correspondence image regional area；
C. according to the filter scale factor p of each regional area of above-mentioned calculating gainedx, calculate the gaussian filtering parameter σ of corresponding regionx
Value be：
<mrow>
<msub>
<mi>&sigma;</mi>
<mi>x</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>p</mi>
<mi>x</mi>
</msub>
<mn>3</mn>
</mfrac>
</mrow>
6. image object extraction system in a kind of water according to claim 5, it is characterized in that, the edge treated unit is used
Handled in filtered image, remove remaining pseudo-edge in filtered image, specifically include：
A. the Grad of filtered image is calculated, so as to set up the histogram of gradients of filtered image, is defined in histogram of gradients
The most gradient magnitude of pixel number is Tm, define the high threshold y of edge connectionm, then threshold value ymCalculation formula be：
<mrow>
<msub>
<mi>k</mi>
<mi>m</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>N</mi>
</msubsup>
<mo>|</mo>
<msub>
<mi>T</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>T</mi>
<mi>m</mi>
</msub>
<mo>|</mo>
</mrow>
<mi>N</mi>
</mfrac>
</mrow>
<mrow>
<msub>
<mi>y</mi>
<mi>m</mi>
</msub>
<mo>=</mo>
<msub>
<mi>T</mi>
<mi>m</mi>
</msub>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>e</mi>
<msub>
<mi>k</mi>
<mi>m</mi>
</msub>
</msup>
<mo>)</mo>
</mrow>
</mrow>
In formula, TiFor the gradient magnitude of image, TmFor the most gradient magnitude of pixel number in histogram of gradients, N is ladder in image
Spend the number of amplitude；
B. remove and be more than above-mentioned gained high threshold y in histogram of gradientsmPixel, set up the histogram of gradients of residual pixel point,
It is K to define the most gradient magnitude of pixel number in the histogram of gradients of residual pixel pointm, define the Low threshold y of edge connectionl,
Then Low threshold ylCalculation formula be：
<mrow>
<msub>
<mi>g</mi>
<mi>l</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>M</mi>
</msubsup>
<mo>|</mo>
<msub>
<mi>K</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>K</mi>
<mi>m</mi>
</msub>
<mo>|</mo>
</mrow>
<mi>M</mi>
</mfrac>
</mrow>
<mrow>
<msub>
<mi>y</mi>
<mi>l</mi>
</msub>
<mo>=</mo>
<msub>
<mi>K</mi>
<mi>m</mi>
</msub>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>e</mi>
<msub>
<mi>g</mi>
<mi>l</mi>
</msub>
</msup>
<mo>)</mo>
</mrow>
</mrow>
In formula, KiFor the Grad in the histogram of gradients of residual pixel point, M is gradient in the histogram of gradients of residual pixel point
The number of amplitude, KmFor the most gradient magnitude of pixel number in the histogram of gradients of residual pixel point；
C. remove and be more than threshold value y in histogram of gradientslPixel, remaining pixel is the edge of image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710424076.0A CN107274391A (en) | 2017-06-07 | 2017-06-07 | A kind of underwater picture Objective extraction system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710424076.0A CN107274391A (en) | 2017-06-07 | 2017-06-07 | A kind of underwater picture Objective extraction system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107274391A true CN107274391A (en) | 2017-10-20 |
Family
ID=60066555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710424076.0A Withdrawn CN107274391A (en) | 2017-06-07 | 2017-06-07 | A kind of underwater picture Objective extraction system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107274391A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108540771A (en) * | 2018-04-02 | 2018-09-14 | 深圳智达机械技术有限公司 | A kind of efficient resource detection system |
| CN111246158A (en) * | 2019-04-15 | 2020-06-05 | 桑尼环保（江苏）有限公司 | Command distribution execution method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020947A (en) * | 2011-09-23 | 2013-04-03 | 阿里巴巴集团控股有限公司 | Image quality analysis method and device |
| CN104567820A (en) * | 2015-01-24 | 2015-04-29 | 无锡桑尼安科技有限公司 | Underwater target central position searching system |
| CN104700421A (en) * | 2015-03-27 | 2015-06-10 | 中国科学院光电技术研究所 | Edge detection algorithm based on canny self-adaptive threshold value |
| CN104999164A (en) * | 2015-08-04 | 2015-10-28 | 李小春 | Underwater robot based on multiple filtering processing |
| CN105070094A (en) * | 2015-08-27 | 2015-11-18 | 上海仪电电子股份有限公司 | Machine vision based parking space detection system and parking space detection method |
-
2017
- 2017-06-07 CN CN201710424076.0A patent/CN107274391A/en not_active Withdrawn
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020947A (en) * | 2011-09-23 | 2013-04-03 | 阿里巴巴集团控股有限公司 | Image quality analysis method and device |
| CN104567820A (en) * | 2015-01-24 | 2015-04-29 | 无锡桑尼安科技有限公司 | Underwater target central position searching system |
| CN104700421A (en) * | 2015-03-27 | 2015-06-10 | 中国科学院光电技术研究所 | Edge detection algorithm based on canny self-adaptive threshold value |
| CN104999164A (en) * | 2015-08-04 | 2015-10-28 | 李小春 | Underwater robot based on multiple filtering processing |
| CN105070094A (en) * | 2015-08-27 | 2015-11-18 | 上海仪电电子股份有限公司 | Machine vision based parking space detection system and parking space detection method |
Non-Patent Citations (2)
| Title |
|---|
| WANG ZHI ET AL.: "Fast Adaptive Threshold for the Canny Edge Detector", 《PROC. OF SPIE》 * |
| 张巍: "水下图像的目标检测与定位研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108540771A (en) * | 2018-04-02 | 2018-09-14 | 深圳智达机械技术有限公司 | A kind of efficient resource detection system |
| CN111246158A (en) * | 2019-04-15 | 2020-06-05 | 桑尼环保（江苏）有限公司 | Command distribution execution method |
| CN111246158B (en) * | 2019-04-15 | 2020-11-06 | 杨丽 | Command distribution execution method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104134209B (en) | A kind of feature extracting and matching method and system in vision guided navigation | |
| CN103824070B (en) | A kind of rapid pedestrian detection method based on computer vision | |
| CN107133973B (en) | Ship detection method in bridge collision avoidance system | |
| CN109800735A (en) | Accurate detection and segmentation method for ship target | |
| CN110275153A (en) | A kind of waterborne target detection and tracking based on laser radar | |
| CN110084165A (en) | The intelligent recognition and method for early warning of anomalous event under the open scene of power domain based on edge calculations | |
| CN106780560B (en) | Bionic robot fish visual tracking method based on feature fusion particle filtering | |
| Selvakumar et al. | The performance analysis of edge detection algorithms for image processing | |
| CN104899866A (en) | Intelligent infrared small target detection method | |
| WO2018000252A1 (en) | Oceanic background modelling and restraining method and system for high-resolution remote sensing oceanic image | |
| CN109583442A (en) | False detection method of license plate and device based on Line segment detection | |
| CN106845410B (en) | Flame identification method based on deep learning model | |
| CN108564602A (en) | Airplane detection method based on airport remote sensing image | |
| CN108520203A (en) | Multiple target feature extracting method based on fusion adaptive more external surrounding frames and cross pond feature | |
| CN103020959B (en) | Gravity model-based oceanic front information extraction method | |
| CN102903108A (en) | Edge detection method based on underwater image statistical property | |
| CN106874912A (en) | A kind of image object detection method based on improvement LBP operators | |
| CN106650580A (en) | Image processing based goods shelf quick counting method | |
| CN104778707A (en) | Electrolytic capacitor detecting method for improving general Hough transform | |
| CN103020967A (en) | Unmanned aerial vehicle aerial image accurate matching method based on island edge characteristics | |
| CN105608689B (en) | A kind of panoramic mosaic elimination characteristics of image error hiding method and device | |
| CN107274391A (en) | A kind of underwater picture Objective extraction system | |
| CN105139391A (en) | Edge detecting method for traffic image in fog-and-haze weather | |
| CN107862262A (en) | A kind of quick visible images Ship Detection suitable for high altitude surveillance | |
| CN110717454B (en) | Wheel type robot obstacle detection method in stage environment |
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: 20171020 |