CN109166132A - A kind of sidescan-sonar image target identification method of variable initial distance sign function - Google Patents
A kind of sidescan-sonar image target identification method of variable initial distance sign function Download PDFInfo
- Publication number
- CN109166132A CN109166132A CN201810779433.XA CN201810779433A CN109166132A CN 109166132 A CN109166132 A CN 109166132A CN 201810779433 A CN201810779433 A CN 201810779433A CN 109166132 A CN109166132 A CN 109166132A
- Authority
- CN
- China
- Prior art keywords
- target
- circle
- function
- sidescan
- sonar image
- 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
- 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
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
A kind of sidescan-sonar image target identification method of variable initial distance sign function, belongs to the field of target recognition of sonar image.The present invention is that solution CV model is insensitive to initial symbol distance function location and shape, noise is easily considered as target during curve evolvement, the problem of easily causing target to lose with noise gray count target gray mean value, a kind of method that location and shape for changing initial distance sign function are added selects the shape and position fit object signature grey scale value and range of suitable initial symbol distance function.It is an advantage of the invention that accurately finding the position of target when target deviates picture medium position；The objective contour of complete display is extracted under big noise situations in picture；This method is simple and reliable, it is easy to accomplish, calculation amount is small, and accuracy is good, improves the feasibility and practicability of side-scan sonar target identification, has positive effect to the development of the job task etc. under water of autonomous underwater vehicle from now on.
Description
Technical field
The invention belongs to the field of target recognition of sonar image, and in particular to a kind of side of variable initial distance sign function
Sweep sonar image target identification method.
Background technique
Sonar is the effective tool for surveying and drawing submarine geomorphy detecting underwater object.Side scan sonar is a kind of typical Active Acoustic
, also referred to as lateral sonar or submarine geomorphy visit side instrument.The basic functional principle of side scan sonar be using ultrasonic array to
Seabed transmitting has directive property, wide vertical beam angle, narrow horizontal beam angle ultrasonic wave, utilize receiving array to receive seabed
Reflection and scattered wave, are processed into acoustic picture eventually by system.It can show submarine geomorphy, determine the Position Approximate and height of target
Degree, has become a kind of technical tool for being widely used in seabed imaging field now.There is target in the signal that sonar receives
Echo, reverberation and noise.Wherein target echo is signal required for forming sonar image, and reverberation and noise can be to sonar images
Quality have an impact, need to take measures on customs clearance by understanding its generation mechanism to be weakened.
In terms of the contours extract of sidescan-sonar image, scholars have used threshold method, clustering procedure, markov random file
The methods of simulation carries out correlative study.But when acoustic image background is complicated, noise pollution is serious, all kinds of threshold methods, cluster
Method is difficult to obtain ideal profile information.And Markov random field model operand in practical calculating process is larger, in reality
There is certain difficulty in the application of border.In recent years, some scholars conduct a research to acoustic image using level set concept, by evolution
Curved surface or curve are embedded into the curved surface of high-dimension function expression, and image letter is described using the change in topology between curve and plane
Breath, to realize the extraction of objective contour.Wherein, Chan-Vese model, abbreviation C-V model are the bases based on Level Set Theory
In the most notable model of regional activity profile.The some region of gray scale of the model selection target substitutes target gray, i.e., by mesh
Mark gray scale is considered as mean value region, seeks the area grayscale mean value in addition to target as background gray levels, then carries out level set
The evolution of equation solves zero level collection and obtains objective contour curve.This method utilizes the gray average of target and background, can
Gray scale in irregular shape is filtered out well close to the noise of background mean value, and there is good robustness.
During level set movements, to avoid the violent concussion in level set function evolutionary process, usually developing
It is preceding that level set function is initialized as symbolic measurement, it is hereafter regularly that level set function is again initial in evolutionary process
Turn to symbolic measurement.But CV model is insensitive to symbolic measurement location and shape, is easy during curve evolvement
Noise is considered as target and be easy to cause target to lose with noise gray count target gray mean value.Therefore, in the base of C-V model
On plinth, a kind of method for changing the location and shape apart from sign function is added, selects the shape of suitable symbolic measurement
It is fitted mesh signature grey scale value and range with position, objective contour can completely be extracted and filter out noise well and obtain clearly
Clear ground objective contour.
Summary of the invention
The purpose of the present invention is to provide a kind of sidescan-sonar image target identification sides of variable initial distance sign function
Method.The present invention is that solution CV model is insensitive to initial symbol distance function location and shape, is easy during curve evolvement
The problem of being considered as target and be easy to cause target to lose with noise gray count target gray mean value noise, is added a kind of change
The method of the location and shape of initial distance sign function selects the shape and position fitting of suitable initial symbol distance function
Mesh signature grey scale value and range, completely to extract objective contour and filter out noise well.
The object of the present invention is achieved like this:
A kind of sidescan-sonar image target identification method of variable initial distance sign function, comprising:
(1.1) level set function is regularly reinitialized to symbolic measurement simultaneously when developing according to level set function
Judge tested point whether the feature inside symbolic measurement, select to judge the smallest circle of calculation amount as symbolic distance letter
Number；
(1.2) it according to selected symbolic measurement circle, is selected in sidescan-sonar image input by target ash to be extracted
Degree feature to the maximum extent envelope into symbolic measurement circle center location and radius size；
(1.3) C-V model evolution level set function is established；
(1.4) active contour model of C-V model is solved using gradient descent method, obtains the parameter of active contour model
EVOLUTION EQUATION.
The symbolic measurement needs to meet | ▽ d |=1, d (x, y)=0 two condition
Gray scale is most in selection target when noisy gray-value is more shallow than target for the status requirement of the described symbolic measurement circle
Deep to distinguish maximum region with noise gray scale, which does the center of circle of symbolic measurement circle, which does symbol
The radius of distance function circle；Select selection target gray scale shallower maximum with the difference of noise gray scale when noise gray scale is than target depth
Region, the regional center do the center of circle of symbolic measurement circle, which does the radius of symbolic measurement circle.
The range that the initial radius of circle of the symbolic measurement circle requires symbolic measurement circle fixed is no more than mesh
Mark range and by target subject envelope into initial symbol distance function circle in, but be no more than target zone.
The C-V model evolution level set function of establishing is to set objective contour curve as C, then
Wherein, φ is the level set form of C, λ1、λ2, ν and μ be all normal number, c1And c2It is the inside and outside portion of curve respectively
Gray average, Hε(φ) indicates curvilinear inner region, (1-Hε(φ)) indicate curved exterior region, Section 3 and Section 4 difference
Indicate the area of length of a curve and curve enclosing region.
The gradient descent method solves, comprising the following steps:
(6.1)Hε(φ) and δε(φ) is respectively the Heaviside function and Dirac function of regularization, and as ε → 0,
There is δε→δ,Hε→ H, is defined as follows:
(6.2) according to strength condition:
Average gray value is calculated using jump function, is obtained:
Wherein c1And c2It is the gray average in the inside and outside portion of curve respectively, for global amount；
(6.3) energy functional in level set function is minimized, corresponding Euler-Lagrange equation develops are as follows:
(6.4) above formula is integrated, solves the value that t moment φ=0 obtains, solves objective contour curve.
The beneficial effects of the present invention are:
(1) a kind of sidescan-sonar image target identification method of variable initial distance sign function proposed by the present invention can
The position of target is accurately found when target deviates picture medium position；
(2) a kind of sidescan-sonar image target identification method of variable initial distance sign function proposed by the present invention can
There are the objective contours that complete display is extracted under big noise situations in picture；
(3) a kind of sidescan-sonar image target identification method of variable initial distance sign function proposed by the present invention is simple
Reliably, it is easy to accomplish, calculation amount is small, and accuracy is good, the feasibility and practicability of side scan sonar target identification is improved, to from now on
Job task etc. development has positive effect to autonomous underwater vehicle under water.
Figure of description
Fig. 1 is a kind of sidescan-sonar image target identification method operational flowchart of variable initial distance sign function；
Fig. 2 is the sidescan-sonar image containing target to be identified；
Fig. 3 is the big logotype in position of the fixed initial symbol distance function of original C-V model；
Fig. 4 is the contour extraction of objects result schematic diagram of the fixed initial symbol distance function of original C-V model；
Fig. 5 is the schematic diagram changed behind initial symbol distance function position and size according to the present invention；
Fig. 6 is using contour extraction of objects result schematic diagram of the invention.
Specific embodiment
Specific embodiments of the present invention will be further explained with reference to the accompanying drawing:
The present invention is a kind of sonar target extracting method of variable initial distance sign function.The present invention includes proposing to improve
The method of active contour model C-V model, the method for changing initial symbol distance function position, selection initial symbol function covering
The method of size.A kind of sidescan-sonar image target identification side of the variable initial distance sign function based on active profile
Method, which comprises the following steps: for sidescan-sonar image, Level Set Theory is introduced in active contour model；Root
According to the feature of target acoustic shadow in sonar image, sonar image is divided into target area and background area；With target area and background area ash
The difference minimum of mean value is spent as evolution target, i.e. target background area energy function is minimum；According to the active contour model of C-V model
The characteristics of extracting objective contour using average gray value inside and outside profile inhibits sidescan-sonar image to be generated by reverberation and noise exhausted
Most of noise extracts the profile that boundary is clearly demarcated, noise is less；Regularly by level set function when being developed according to level set function
Be reinitialized to symbolic measurement and judge tested point whether the feature inside symbolic measurement, selection judgement calculate
The smallest circle is measured as symbolic measurement, wherein the center location of symbolic measurement circle and radius size are variable；According to choosing
Fixed symbolic measurement circle, in sidescan-sonar image input, selection is by target gray feature to be extracted envelope to the maximum extent
Into the center location and radius size of symbolic measurement circle；The active contour model of C-V model is asked using gradient descent method
Solution, obtains the Parameters Evolution equation of active contour model.
(1) round initial distance sign function
Level set equation is initialized as symbolic measurement when solving objective contour and in level by conventional flat set method
Violent concussion when collection equation is periodically initialized as apart from sign function to avoid the evolution of level set equation during developing is made
It at objective contour confusion or loses, level set function is initialized as symbolic measurement usually before evolution, is hereafter being developed
Level set function is regularly reinitialized to symbolic measurement in the process.But C-V model is to initial symbol distance function
Location and shape are insensitive, during curve evolvement be easy by noise be considered as target and with noise gray count target gray it is equal
Value be easy to cause target to lose.
In order to solve this problem, a kind of method of reindexing distance function is proposed, ideal target is finally obtained
Profile.Due to need to only determine that center location and radius are assured that round size and location selection is convenient and is iterating to calculate
Judge that only needing the selected center of circle to calculate arbitrary point on image and do to the Euclidean distance in the center of circle with radius when the point inside and outside target subtracts
Method can calculation amount be small can greatly shorten the Objective extraction time, so selecting circle as symbolic measurement.According to a kind of base
In the sidescan-sonar image target identification method of the variable initial distance sign function of active profile, select circle as apart from symbol
Function.Position and the radius size of initial SDF circle are selected in initialization, with mouse completely to extract profile and preferable
Ground filters out noise.The range for requiring SDF circle fixed when selection is no more than target zone.According to a kind of variable based on active profile
The sidescan-sonar image target identification method of initial distance sign function selects initial SDF circle position.It is required that working as noisy gray-value
Gray scale is most deep in selection target when more shallow than target distinguishes maximum region with noise gray scale；When noise gray scale when target depth than selecting
Selection target gray scale is shallower to distinguish maximum region with noise gray scale.According to a kind of variable initial distance symbol based on active profile
The sidescan-sonar image target identification method of number function, selects initial SDF radius of circle.It is required that the fixed range of SDF circle is no more than
Target zone and by target subject envelope into initial SDF circle in.But it is no more than target zone against background and noise gray scale
It influences.
Its equation when selection circle does symbolic measurement are as follows:
φ (x, y)=x2+y2-1
Wherein, | ▽ φ |=1 and φ (x, y)=0.This is round implicit function expression-form, the function representation two-dimensional surface
Plane Ω is divided into two regions, Ω by the unit circle on the Ω of space-For the interior zone of unit, meet small φ (x, y) < 0；
Ω+For the perimeter of unit circle, meet φ (x, y) > 0.Unit circular curve is represented by Ω, meets symbolic measurement item
Part: | ▽ φ |=1 and φ (x, y)=0, affiliated three region divisions specifically: when φ (x, y) < 0, indicate symbolic measurement
Round interior zone；When φ (x, y) > 0, the perimeter of symbolic measurement circle is indicated；When φ (x, y)=0, symbol is indicated
Distance function circular curve.Sidescan-sonar image input when, selection by target gray feature to be extracted to the maximum extent envelope into
The center location and radius size of symbolic measurement circle.
(2) the energy function model of C-V model
The energy function of C-V model shows respectively the inside and outside gray variance of contour curve, the final mesh of evolution
Mark is to make the inside and outside variance of curve while reaching minimum.Its energy function formula are as follows:
In formula, λ1、λ2, ν and μ be all normal number, c1And c2, it is the gray average H in the inside and outside portion of curve respectivelyε(φ) is indicated
Curvilinear inner region, (1-Hε(φ)) indicate curved exterior region, Section 3 and Section 4 respectively indicate length of a curve and song
The area of line enclosing region.
(3) solution of C-V active contour model
H in equationε(φ) and δε(φ) is respectively the Heaviside function and Dirac function of regularization, and as ε → 0,
There is δε→δ,Hε→ H, is defined as follows:
Then corresponding level set letter is minimized by solving the corresponding Euler-Lagrange equation of energy functional to realize
Number EVOLUTION EQUATION is formula:
In formula:
c1And c2It is the gray average in the inside and outside portion of curve respectively, for global amount.After testing image input, initial symbol is determined
The size and location of distance function can introduce time t, solve partial differential equation using gradient descent method.
(4) symbolic measurement is adjusted when image inputs
Operation platform is Matlab 2014a, the testing image after input filter denoising.
Operation is shut down procedure using Pause function in Matlab function library in image initial, prevents level set equation
Iteration reads keyboard input numerical value as SDF radius of circle using the center location of mouse selection SDF circle.
The principle of selection is:
When noisy gray-value is more shallow than target, gray scale is most deep in selection target distinguishes in maximum region with noise gray scale
The heart does the center of circle SDF, which does radius.
Similarly, it selects selection target gray scale shallower when noise gray scale is than target depth and noise gray scale distinguishes maximum area
The center of circle SDF is done at the center in domain, which does radius.
The range for requiring SDF circle fixed when selection is no more than target zone.
(5) experimental result
It is verified using method of the Matlab Programming with Pascal Language to text, in order to compare, to provide initial symbol distance function solid
Determining evolution result and initial symbol distance function can be changed and drill experimental result.
Fig. 1 is the sidescan-sonar image that a width contains target to be identified.Fig. 2 is original C-V model initial distance symbol letter
Numerical digit is set and size, and Fig. 3 is result of the original C-V model to sonar image contour extraction of objects.From the results, it was seen that due to
Original C-V model initial distance sign function is fixed on image central part, can not flexibly cope in sonar image different location
The target of appearance carries out flexible initial profile sampled grey, therefore partial target loss and showing by noise jamming occurs
As.
After Fig. 4 is input picture, initial symbol is found apart from letter according to text the method when initializing level set equation
The radius size of numerical digit circle position and setting, Fig. 5 are the results to develop according to change apart from sign function method iteration.From result
As can be seen that methods herein can correctly find target to be identified in sonar image there are in the case where noise, have very
Good noise immunity, and extract clear, complete objective contour.
(1) select circle as apart from sign function；In initialization with mouse selection initial symbol distance function (SDF) circle
Position and radius size, completely to extract profile and preferably filter out noise.The range for requiring SDF circle fixed when selection
No more than target zone.
The advantages of generally selecting round symbolic measurement, doing so is only need to determine that center location and radius are assured that
Round size and location selection is convenient and is only needed on selected center of circle calculating figure when iterative calculation judges the point inside and outside target
The Euclidean distance and radius in arbitrary point to the center of circle do that subtraction can calculation amount be small can greatly shorten the Objective extraction time.
Select to need to meet when symbolic measurement | ▽ d |=1, d (x, y)=0 two condition, it was demonstrated that process is as follows.
For a plane Ω, a symbolic measurement φ is defined, and plane Ω is divided into two regions by the function.
It is calculated are as follows:
φ (x, y, t=0)=Sign (x, y) d (x, y) (1)
Wherein, Sign () characterizes the symbol of distance, and even point is in curvilinear inner, Sign ()=- 1, conversely, in curve
Outside is Sign ()=+ 1；Distance function d (x, y) be point (x, y) to curve C (x, y) shortest distance Euclidean distance into
Row calculates.
It is indicated apart from sign function d (v) are as follows:
D (v)=min | v-vi|},vi∈Ω (2)
For boundaryOn point ν meet d (ν)=0, otherwise for appoint be not intended to borderline point ν, if on boundary
On find the point ν nearest with νc, then d (ν)=| ν-νc|.According to Euclidean distance, d (ν) expression are as follows:
The then gradient of d are as follows:
Known by above formula | ▽ d |=1.The operation that reinitializes periodically carried out during curve evolvement is according to zero level collection
The closed curve C (x, y, t) of expression recalculates symbolic measurement, to substitute current level set function φ (x, y, t).By
Meet zero level set function in symbolic measurement, requires φ (x, y)=0 when it requires t=0, i.e. d (x, y)=0.So
Select to need to meet when symbolic measurement | ▽ d |=1, d (x, y)=0 two condition.
Circle symbolic measurement meets the condition apart from sign function, selects it as symbolic measurement, authenticated
Journey is as follows.
If y=f (x) be any one curve of two-dimensional surface spatially, enable any point on the function representation curve (x,
Y) a kind of corresponding relationship of abscissa x and ordinate y, the expression of this corresponding relationship are then y-f (x)=0.If setting:
φ (x, y)=y-f (x, y) (5)
Then φ (x, y)=0 is the implicit expression formula of two-dimensional surface space curve.Assuming that the level on the Ω of two-dimensional surface space
Set function are as follows:
φ (x, y)=x2+y2-1 (6)
Plane Ω is divided into two regions, Ω by the unit circle on function representation two-dimensional surface space Ω-For unit
Interior zone meets small φ (x, y) < 0；Ω+For the perimeter of unit circle, meet φ (x, y) > 0；Unit circle curve indicates
For Ω, meet φ (x, y)=0
Therefore the equation of circle meets symbolic measurement | ▽ φ | the condition of=1 and φ (x, y)=0.
(2) selection initial symbol distance circle position and radius；Select initial SDF circle status requirement when noisy gray-value compares mesh
Gray scale is most deep in selection target when marking shallow distinguishes maximum region with noise gray scale, which does the center of circle SDF, and the region is big
It is small to do radius；Select selection target gray scale is shallower to distinguish maximum region with noise gray scale when noise gray scale is than target depth, it should
Regional center does the center of circle SDF, which does radius；Selecting initial SDF radius of circle requirement SDF circle, fixed range is no more than
Target zone and by target subject envelope into initial SDF circle in, but be no more than target zone against background and noise gray scale
It influences.
(3) C-V model evolution level set function is established；If objective contour curve is C, then have:
Wherein, φ is the level set form of C, λ1、λ2, ν and μ be all normal number, c1And c2It is the inside and outside portion of curve respectively
Gray average, Hε(φ) indicates curvilinear inner region, (1-Hε(φ)) indicate curved exterior region, Section 3 and Section 4 difference
Indicate the area of length of a curve and curve enclosing region.
(4) C-V active contour model is solved, the active contour model of C-V model is solved using gradient descent method, is obtained
The Parameters Evolution equation of active contour model, comprising the following steps:
Step 1: H in equation (7)ε(φ) and δε(φ) is respectively the Heaviside function and Dirac function of regularization,
And as ε → 0, there is δε→δ,Hε→ H, is defined as follows:
Step 2: according to strength condition:
Average gray value is calculated using jump function, is obtained:
Wherein c1And c2It is the gray average in the inside and outside portion of curve respectively, for global amount.
Step 3: we can minimize the energy functional in formula (7) by energy function obtained above.Its is right
The Euler-Lagrange equation answered develops are as follows:
This is the partial differential equation that the level set function of (7) formula obtains time parameter derivation.
Step 4: being integrated to (13) formula, the value that t moment φ=0 obtains is solved, objective contour curve can be solved.
Claims (6)
1. a kind of sidescan-sonar image target identification method of variable initial distance sign function characterized by comprising
(1.1) level set function is regularly reinitialized to symbolic measurement when developing according to level set function and is judged
Tested point whether the feature inside symbolic measurement, select to judge the smallest circle of calculation amount as symbolic measurement；
(1.2) according to selected symbolic measurement circle, in sidescan-sonar image input, selection is special by target gray to be extracted
Point to the maximum extent envelope into symbolic measurement circle center location and radius size；
(1.3) C-V model evolution level set function is established；
(1.4) active contour model of C-V model is solved using gradient descent method, obtains the Parameters Evolution of active contour model
Equation.
2. a kind of sidescan-sonar image target identification method of variable initial distance sign function according to claim 1,
It is characterized in that, the symbolic measurement needs to meet | ▽ d |=1, d (x, y)=0 two condition.
3. a kind of sidescan-sonar image target identification method of variable initial distance sign function according to claim 1,
It is characterized by: the status requirement of described symbolic measurement circle gray scale in selection target when noisy gray-value is more shallow than target
Most deep to distinguish maximum region with noise gray scale, which does the center of circle of symbolic measurement circle, which accords with
The radius of number distance function circle；Select selection target gray scale shallower maximum with the difference of noise gray scale when noise gray scale is than target depth
Region, the regional center do symbolic measurement circle the center of circle, the area size do symbolic measurement circle radius.
4. a kind of sidescan-sonar image target identification method of variable initial distance sign function according to claim 4,
It is characterized by: the initial radius of circle of the symbolic measurement circle requires symbolic measurement circle, fixed range is no more than
Target zone and by target subject envelope into initial symbol distance function circle in, but be no more than target zone.
5. a kind of sidescan-sonar image target identification method of variable initial distance sign function according to claim 1,
It is characterized by: the C-V model evolution level set function of establishing is to set objective contour curve as C, then
Wherein, φ is the level set form of C, λ1、λ2, ν and μ be all normal number, c1And c2Be respectively the inside and outside portion of curve gray scale it is equal
Value, Hε(φ) indicates curvilinear inner region, (1-Hε(φ)) indicate curved exterior region, Section 3 and Section 4 respectively indicate song
The length of line and the area of curve enclosing region.
6. a kind of sidescan-sonar image target identification method of variable initial distance sign function according to claim 1,
It is characterized by: the gradient descent method solves, comprising the following steps:
(6.1)Hε(φ) and δε(φ) is respectively the Heaviside function and Dirac function of regularization, and as ε → 0, there is δε
→δ,Hε→ H, is defined as follows:
(6.2) according to strength condition:
Average gray value is calculated using jump function, is obtained:
Wherein c1And c2It is the gray average in the inside and outside portion of curve respectively, for global amount；
(6.3) energy functional in level set function is minimized, corresponding Euler-Lagrange equation develops are as follows:
(6.4) above formula is integrated, solves the value that t moment φ=0 obtains, solves objective contour curve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810779433.XA CN109166132B (en) | 2018-07-16 | 2018-07-16 | Side-scan sonar image target identification method with variable initial distance symbolic function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810779433.XA CN109166132B (en) | 2018-07-16 | 2018-07-16 | Side-scan sonar image target identification method with variable initial distance symbolic function |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109166132A true CN109166132A (en) | 2019-01-08 |
| CN109166132B CN109166132B (en) | 2022-01-07 |
Family
ID=64897843
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810779433.XA Active CN109166132B (en) | 2018-07-16 | 2018-07-16 | Side-scan sonar image target identification method with variable initial distance symbolic function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109166132B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110048783A (en) * | 2019-04-19 | 2019-07-23 | 美钻深海能源科技研发（上海）有限公司 | Sound wave communication unit, system, base station, inspection device and sound wave communication method |
| CN111260674A (en) * | 2020-01-14 | 2020-06-09 | 武汉理工大学 | Method, system and storage medium for extracting target contour line from sonar image |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779346A (en) * | 2012-07-05 | 2012-11-14 | 西安电子科技大学 | SAR (storage address register) image changing detection method based on improved C-V model |
| CN103886599A (en) * | 2014-03-26 | 2014-06-25 | 北京工业大学 | Blood vessel ROI dividing method based on intravascular ultrasonic image |
| CN104217422A (en) * | 2014-06-03 | 2014-12-17 | 哈尔滨工程大学 | Sonar image detection method of self-adaption narrow-band level set |
| CN104463889A (en) * | 2014-12-19 | 2015-03-25 | 中国人民解放军国防科学技术大学 | Unmanned plane autonomous landing target extracting method based on CV model |
| CN104657736A (en) * | 2013-11-19 | 2015-05-27 | 中国科学院沈阳自动化研究所 | Active contour-based sonar image mine target recognition method |
| CN105469408A (en) * | 2015-11-30 | 2016-04-06 | 东南大学 | Building group segmentation method for SAR image |
| CN107240108A (en) * | 2017-06-06 | 2017-10-10 | 衢州学院 | Movable contour model image partition method based on local Gaussian fitting of distribution and the poor energy driving of local symbol |
-
2018
- 2018-07-16 CN CN201810779433.XA patent/CN109166132B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779346A (en) * | 2012-07-05 | 2012-11-14 | 西安电子科技大学 | SAR (storage address register) image changing detection method based on improved C-V model |
| CN104657736A (en) * | 2013-11-19 | 2015-05-27 | 中国科学院沈阳自动化研究所 | Active contour-based sonar image mine target recognition method |
| CN103886599A (en) * | 2014-03-26 | 2014-06-25 | 北京工业大学 | Blood vessel ROI dividing method based on intravascular ultrasonic image |
| CN104217422A (en) * | 2014-06-03 | 2014-12-17 | 哈尔滨工程大学 | Sonar image detection method of self-adaption narrow-band level set |
| CN104463889A (en) * | 2014-12-19 | 2015-03-25 | 中国人民解放军国防科学技术大学 | Unmanned plane autonomous landing target extracting method based on CV model |
| CN105469408A (en) * | 2015-11-30 | 2016-04-06 | 东南大学 | Building group segmentation method for SAR image |
| CN107240108A (en) * | 2017-06-06 | 2017-10-10 | 衢州学院 | Movable contour model image partition method based on local Gaussian fitting of distribution and the poor energy driving of local symbol |
Non-Patent Citations (3)
| Title |
|---|
| CHUNMING LI等: "Distance Regularized Level Set Evolution and Its Application to Image Segmentation", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
| 唐利明等: "结合BV-L2分解的CV变分水平集模型", 《重庆大学学报》 * |
| 官俊等: "基于圆形约束CV-LIF模型的原木端面图像分割", 《计算机工程与应用》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110048783A (en) * | 2019-04-19 | 2019-07-23 | 美钻深海能源科技研发（上海）有限公司 | Sound wave communication unit, system, base station, inspection device and sound wave communication method |
| CN111260674A (en) * | 2020-01-14 | 2020-06-09 | 武汉理工大学 | Method, system and storage medium for extracting target contour line from sonar image |
| CN111260674B (en) * | 2020-01-14 | 2023-04-18 | 武汉理工大学 | Method, system and storage medium for extracting target contour line from sonar image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109166132B (en) | 2022-01-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Reed et al. | An automatic approach to the detection and extraction of mine features in sidescan sonar | |
| Luo et al. | Sediment classification of small-size seabed acoustic images using convolutional neural networks | |
| CN109239719B (en) | Multi-beam forward-looking sonar obstacle area extraction method integrating multi-frame information | |
| US10497128B2 (en) | Method and system for sea background modeling and suppression on high-resolution remote sensing sea images | |
| CN107167810B (en) | Method for rapidly extracting underwater target by side-scan sonar imaging | |
| CN104217422B (en) | A kind of sonar image detection method of adaptive narrow-band level set | |
| CN105182328A (en) | Ground penetrating radar underground target positioning method | |
| CN104049245A (en) | Urban building change detection method based on LiDAR point cloud spatial difference analysis | |
| EP2757529B1 (en) | Systems and methods for 3D data based navigation using descriptor vectors | |
| CN109166132A (en) | A kind of sidescan-sonar image target identification method of variable initial distance sign function | |
| Maussang et al. | Mean–standard deviation representation of sonar images for echo detection: Application to SAS images | |
| Kohntopp et al. | Seafloor classification for mine countermeasures operations using synthetic aperture sonar images | |
| CN106815851B (en) | A kind of grid circle oil level indicator automatic reading method of view-based access control model measurement | |
| CN107292920A (en) | A kind of multidate full-polarization SAR remote sensing imagery change detection method of joint classification | |
| Long et al. | A comprehensive deep learning-based outlier removal method for multibeam bathymetric point cloud | |
| CN112907615B (en) | Submarine landform unit contour and detail identification method based on region growing | |
| EP2757526B1 (en) | Systems and methods for 3D data based navigation using a watershed method | |
| Padial et al. | Correlation of imaging sonar acoustic shadows and bathymetry for ROV terrain-relative localization | |
| CN106780332B (en) | Whole-borehole logging image generation device | |
| CN106600535B (en) | Whole-borehole logging image generation method | |
| WO2022195675A1 (en) | Information processing device, control method, program, and storage medium | |
| CN102842139B (en) | A kind of acquiring method of target trajectory | |
| CN108051811B (en) | Method and device for analyzing mesoscale vortex | |
| Hao et al. | Data fusion of multi-beam sonar and side-scan sonar base on feature contour registration | |
| Febriawan et al. | Detection and characterization of an archaeological wreck site in Sunda Strait, Indonesia |
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 |