CN112766287A - SAR image ship target detection acceleration method based on density examination - Google Patents

SAR image ship target detection acceleration method based on density examination Download PDF

Info

Publication number
CN112766287A
CN112766287A CN202110161176.5A CN202110161176A CN112766287A CN 112766287 A CN112766287 A CN 112766287A CN 202110161176 A CN202110161176 A CN 202110161176A CN 112766287 A CN112766287 A CN 112766287A
Authority
CN
China
Prior art keywords
super
density
pixel
clutter
sar 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
Application number
CN202110161176.5A
Other languages
Chinese (zh)
Other versions
CN112766287B (en
Inventor
李刚
王学谦
刘瑜
何友
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
Naval Aeronautical University
Original Assignee
Tsinghua University
Naval Aeronautical University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tsinghua University, Naval Aeronautical University filed Critical Tsinghua University
Priority to CN202110161176.5A priority Critical patent/CN112766287B/en
Publication of CN112766287A publication Critical patent/CN112766287A/en
Application granted granted Critical
Publication of CN112766287B publication Critical patent/CN112766287B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications
    • G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G01S13/9021—SAR image post-processing techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • 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)
  • Theoretical Computer Science (AREA)
  • Remote Sensing (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Multimedia (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Probability & Statistics with Applications (AREA)
  • Electromagnetism (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention provides a density examination-based SAR image ship target detection acceleration method, and belongs to the field of synthetic aperture radar image processing. Firstly, performing superpixel segmentation on an SAR image; respectively calculating the density characteristic and the density distance characteristic of each super pixel, and screening out the super pixels of the SAR image serving as a target clustering center and the super pixels serving as a clutter clustering center; and comparing the difference between each super pixel and the clutter clustering center and the difference between the super pixel and the target clustering center by using a nearest neighbor classifier, deleting the super pixel with larger difference with the target clustering center, and taking the finally reserved super pixel as the input of a subsequent ship target detection method. According to the method, a large number of sea clutter superpixels are deleted quickly before detection, the reserved superpixels are used as the input of the ship target detection method and fine detection is carried out, the calculation speed of the existing superpixel detection method is obviously improved, and the running efficiency of ship target detection in the SAR image is improved.

Description

SAR image ship target detection acceleration method based on density examination
Technical Field
The invention belongs to the field of Synthetic Aperture Radar (SAR) image processing, and particularly relates to a SAR image ship target detection acceleration method based on density examination, which can be used for quickly realizing the conventional superpixel ship target detection method.
Background
Synthetic Aperture Radar (SAR) is an active imaging device that can provide high resolution imaging results of sea surface vessel targets. Compared with optical and infrared sensors, SAR imaging is hardly affected by illumination and weather, and the SAR imaging sensor is a sensor with all-weather and all-time working capability. The ship target detection in the SAR image has important application in the aspects of military sea defense, civil ship monitoring, sustainable fishery and the like.
In recent years, a plurality of expert scholars propose various superpixel detection methods, and good effects are obtained in SAR image ship target detection. However, the existing superpixel detection method needs sliding window fine processing of each superpixel in the SAR image, and the running speed is very slow, so that the calculation efficiency of the detection process is greatly reduced.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a SAR image ship target detection acceleration method based on density examination. According to the method, a large number of sea clutter superpixels are deleted quickly before detection, the reserved superpixels are used as input of a subsequent ship target detection method, the reserved superpixels are subjected to fine detection, the calculation speed of the conventional superpixels detection method is obviously improved, and the running efficiency of ship target detection in the SAR image is improved.
The invention provides a density review-based SAR image ship target detection acceleration method which is characterized in that the method comprises the steps of firstly carrying out superpixel segmentation on an SAR image; respectively calculating the density characteristic and the density distance characteristic of each super pixel, and screening out the super pixels of the SAR image serving as a target clustering center and the super pixels serving as a clutter clustering center; and comparing the difference between each super pixel and the clutter clustering center and the difference between the super pixel and the target clustering center by using a nearest neighbor classifier, deleting the super pixel with larger difference with the target clustering center, and taking the finally reserved super pixel as the input of a subsequent ship target detection method. The method comprises the following steps:
1) obtaining an SAR image, wherein the number of pixels of the image is N; setting the super pixel size S, the number of super pixels in the image is
Figure BDA0002936728400000011
Represents rounding up; setting a regularization parameter lambda to be more than 0;
2) super-pixel segmentation;
taking the size S of the super-pixel, a regularization parameter lambda and the SAR image as input, and obtaining all the super-pixels in the SAR image by using a simple linear iterative clustering SLIC algorithm;
3) calculating the density characteristic rho of each super-pixeli：
Figure BDA0002936728400000021
Wherein I is 1,2, …, and I, I is the index of the super pixel; j denotes an index of a pixel other than the ith pixel in the image, Di,j＝|μi-μjI represents the gray mean value mu of the ith super pixeliAnd the jth super pixel mean value mujMu represents the super-pixel gray-scale mean value,
Figure BDA0002936728400000022
representing a soft truncation distance, alpha is an input scale factor, and alpha belongs to (0, 1);
4) calculating a density distance feature r for each superpixeli：
Figure BDA0002936728400000023
Wherein, gamma isi＝{j|ρj＜ρiJ 1,2, I, j ≠ I } represents a set of superpixels that are less dense than the ith superpixel,
Figure BDA0002936728400000024
representing an empty set;
5) respectively calculating a target clustering center and a clutter clustering center of the SAR image;
firstly, the density characteristics rhoiAnd each density distance characteristic riNormalized to the interval [0,1 ] respectively]To obtain normalized density features
Figure BDA0002936728400000025
And normalized density distance features
Figure BDA0002936728400000026
Reusing normalized density features
Figure BDA0002936728400000027
And normalized density distance features
Figure BDA0002936728400000028
Finding the image target clustering center iTargetClustering center of sum clutter iClutter：
Figure BDA0002936728400000029
Figure BDA00029367284000000210
Wherein, CTargetRepresenting the superpixel index as the center of the target cluster among all superpixels, CClutterRepresenting a superpixel index as a clutter cluster center among all superpixels;
6) deleting clutter superpixels in the SAR image by using a nearest neighbor classifier;
for each super-pixel, the following is determined:
Figure BDA0002936728400000031
wherein,
Figure BDA0002936728400000032
representing the difference between the ith superpixel and the clutter cluster center;
Figure BDA0002936728400000033
representing the difference between the current superpixel and the target cluster center;
Figure BDA0002936728400000034
representing the density feature value of the superpixel in the image as the clutter cluster center,
Figure BDA0002936728400000035
representing the density characteristic value of the superpixel in the image as the center of the target cluster,
Figure BDA0002936728400000036
representing the density distance eigenvalues of superpixels in the image that are the clutter cluster centers,
Figure BDA0002936728400000037
representing a density distance characteristic value of a super pixel serving as a target clustering center in the image;
7) outputting the super pixels reserved after the step 6) in the SAR image.
The invention has the characteristics and beneficial effects that:
the conventional SAR image superpixel ship target detection method needs to finely detect each superpixel in an image, and the fine detection of each superpixel usually needs complex operations such as matrix inversion, parameter iterative estimation and the like, so that the calculation operation efficiency of the superpixel detection method is low. The invention provides a SAR image ship target detection accelerating method based on density examination based on the density characteristics of an SAR image.
Drawings
FIG. 1 is an overall flow chart of the method of the present invention.
Detailed Description
The invention provides a density censored-based SAR image ship target detection acceleration method, which is further described in detail below by combining the accompanying drawings and specific embodiments.
The invention provides a density review-based SAR image ship target detection acceleration method, the overall process is shown in figure 1, and the method comprises the following steps:
1) acquiring an SAR image, where the number of pixels of the image is N (for example, N is 3000 × 3000); setting a superpixel size S (S can be set to be 25% of the number of pixels occupied by ships, S is generally 10-100, and S is 30 in the embodiment), and then the number of superpixels in the image is
Figure BDA0002936728400000041
Represents rounding up; the regularization parameter λ > 0 is set for superpixel segmentation (in this embodiment, the parameter λ is 0.4).
2) Super-pixel segmentation;
taking the size S of the super-pixel, the regularization parameter lambda and the SAR image as input, and obtaining all the super-pixels in the SAR image according to a Simple Linear Iterative Clustering (SLIC) algorithm in the Ship Detection With super-Level Fisher Vector in High-Resolution SAR Images published by the IEEE Geoscience and Remote Sensing rules.
3) Calculating the density characteristic rho of each super-pixeli；
Figure BDA0002936728400000042
Wherein I is 1,2, …, I is the index of the superpixel, and I is the number of the superpixels in the SAR image; j denotes an index of a pixel other than the ith pixel in the image, Di,j＝|μi-μjI represents the gray mean value mu of the ith super pixeliAnd the jth super pixel mean value mujMu represents the super-pixel gray-scale mean value,
Figure BDA0002936728400000043
the soft truncation distance is represented by a constant, α is an input scale factor, and the scale factor α ∈ (0,1) is used to calculate the density feature, where α is 0.3.
4) Calculating a density distance feature r for each superpixeli：
Figure BDA0002936728400000044
Wherein, gamma isi＝{j|ρj＜ρiJ 1, 2.. I, j ≠ I } represents a set of superpixels that is lower in density than the ith superpixel.
Figure BDA0002936728400000045
Indicating an empty set.
5) And respectively calculating a target clustering center and a clutter clustering center of the SAR image.
Firstly, the density characteristics rhoiAnd each density distance characteristic riNormalized to the interval [0,1 ] respectively]To obtain normalized density features
Figure BDA0002936728400000046
And normalized density distance features
Figure BDA0002936728400000047
The normalization can be performed as follows:
Figure BDA0002936728400000048
reusing normalized density features
Figure BDA0002936728400000051
And normalized density distance features
Figure BDA0002936728400000052
Finding the image target clustering center iTargetClustering center of sum clutter iClutter：
Figure BDA0002936728400000053
Figure BDA0002936728400000054
Wherein, CTargetRepresenting the superpixel index as the center of the target cluster among all superpixels, CClutterRepresenting the superpixel index of all superpixels as the clutter cluster center.
6) Deleting clutter superpixels in the SAR image by using a nearest neighbor classifier;
for each super-pixel, the following is determined:
Figure BDA0002936728400000055
wherein
Figure BDA0002936728400000056
Representing the difference between the ith superpixel and the clutter cluster center,
Figure BDA0002936728400000057
representing the difference of the current superpixel from the target cluster center.
Figure BDA0002936728400000058
Representing density features of superpixels as clutter cluster centers in an imageThe value of the one or more of,
Figure BDA0002936728400000059
representing the density characteristic value of the superpixel in the image as the center of the target cluster,
Figure BDA00029367284000000510
representing the density distance eigenvalues of superpixels in the image that are the clutter cluster centers,
Figure BDA00029367284000000511
representing the density distance eigenvalue of the superpixel in the image as the center of the target cluster.
7) Outputting the super pixels (which are regarded as the candidates of the target super pixels) reserved after the step 6) in the SAR image.
In the invention, because formula (5) deletes a large amount of clutter superpixels, the running time of the subsequent superpixel detection method is obviously reduced.
The retained superpixels output in the step 7) can be directly used as the input of the conventional SAR image superpixel ship target detection method, namely, the conventional SAR image superpixel ship target detection method does not need to perform fine detection on all superpixels in the SAR image, and only needs to perform fine detection on a small number of retained superpixels output in the step 7), so that the time required for completing the detection is greatly reduced, and the calculation operation efficiency is improved.

Claims (3)

1. A SAR image ship target detection acceleration method based on density examination is characterized in that the method firstly carries out superpixel segmentation on an SAR image; respectively calculating the density characteristic and the density distance characteristic of each super pixel, and screening out the super pixels of the SAR image serving as a target clustering center and the super pixels serving as a clutter clustering center; and comparing the difference between each super pixel and the clutter clustering center and the difference between the super pixel and the target clustering center by using a nearest neighbor classifier, deleting the super pixel with larger difference with the target clustering center, and taking the finally reserved super pixel as the input of a subsequent ship target detection method.
2. A method as claimed in claim 1, characterized in that the method comprises the following steps:
1) obtaining an SAR image, wherein the number of pixels of the image is N; setting the super pixel size S, the number of super pixels in the image is
Figure FDA0002936728390000011
Figure FDA0002936728390000012
Represents rounding up; setting a regularization parameter lambda to be more than 0;
2) super-pixel segmentation;
taking the size S of the super-pixel, a regularization parameter lambda and the SAR image as input, and obtaining all the super-pixels in the SAR image by using a simple linear iterative clustering SLIC algorithm;
3) calculating the density characteristic rho of each super-pixeli：
Figure FDA0002936728390000013
Wherein I is 1,2, …, and I, I is the index of the super pixel; j denotes an index of a pixel other than the ith pixel in the image, Di,j＝|μi-μjI represents the gray mean value mu of the ith super pixeliAnd the jth super pixel mean value mujMu represents the super-pixel gray-scale mean value,
Figure FDA0002936728390000014
representing a soft truncation distance, alpha is an input scale factor, and alpha belongs to (0, 1);
4) calculating a density distance feature r for each superpixeli：
Figure FDA0002936728390000015
Wherein, gamma isi＝{j|ρj＜ρiJ 1,2, I, j ≠ I } represents a set of superpixels that are less dense than the ith superpixel,
Figure FDA0002936728390000016
representing an empty set;
5) respectively calculating a target clustering center and a clutter clustering center of the SAR image;
firstly, the density characteristics rhoiAnd each density distance characteristic riNormalized to the interval [0,1 ] respectively]To obtain normalized density features
Figure FDA0002936728390000021
And normalized density distance features
Figure FDA0002936728390000022
Reusing normalized density features
Figure FDA0002936728390000023
And normalized density distance features
Figure FDA0002936728390000024
Finding the image target clustering center iTargetClustering center of sum clutter iClutter：
Figure FDA0002936728390000025
Figure FDA0002936728390000026
Wherein, CTargetRepresenting the superpixel index as the center of the target cluster among all superpixels, CClutterRepresenting a superpixel index as a clutter cluster center among all superpixels;
6) deleting clutter superpixels in the SAR image by using a nearest neighbor classifier;
for each super-pixel, the following is determined:
Figure FDA0002936728390000027
wherein,
Figure FDA0002936728390000028
representing the difference between the ith superpixel and the clutter cluster center;
Figure FDA0002936728390000029
representing the difference between the current superpixel and the target cluster center;
Figure FDA00029367283900000210
representing the density feature value of the superpixel in the image as the clutter cluster center,
Figure FDA00029367283900000211
representing the density characteristic value of the superpixel in the image as the center of the target cluster,
Figure FDA00029367283900000212
representing the density distance eigenvalues of superpixels in the image that are the clutter cluster centers,
Figure FDA00029367283900000213
representing a density distance characteristic value of a super pixel serving as a target clustering center in the image;
7) outputting the super pixels reserved after the step 6) in the SAR image.
3. The method of claim 2, wherein the normalization method in step 5) is as follows:
Figure FDA00029367283900000214
CN202110161176.5A 2021-02-05 2021-02-05 SAR image ship target detection acceleration method based on density examination Active CN112766287B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110161176.5A CN112766287B (en) 2021-02-05 2021-02-05 SAR image ship target detection acceleration method based on density examination

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110161176.5A CN112766287B (en) 2021-02-05 2021-02-05 SAR image ship target detection acceleration method based on density examination

Publications (2)

Publication Number Publication Date
CN112766287A true CN112766287A (en) 2021-05-07
CN112766287B CN112766287B (en) 2021-09-17

Family

ID=75705159

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110161176.5A Active CN112766287B (en) 2021-02-05 2021-02-05 SAR image ship target detection acceleration method based on density examination

Country Status (1)

Country Link
CN (1) CN112766287B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9389311B1 (en) * 2015-02-19 2016-07-12 Sandia Corporation Superpixel edges for boundary detection
CN107067039A (en) * 2017-04-25 2017-08-18 西安电子科技大学 SAR image Ship Target quick determination method based on super-pixel
CN108629783A (en) * 2018-05-02 2018-10-09 山东师范大学 Image partition method, system and medium based on the search of characteristics of image density peaks
CN108765491A (en) * 2018-05-31 2018-11-06 成都信息工程大学 A kind of SAR image Ship Target Detection method
CN109886218A (en) * 2019-02-26 2019-06-14 西安电子科技大学 SAR image Ship Target Detection method based on super-pixel statistics diversity

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9389311B1 (en) * 2015-02-19 2016-07-12 Sandia Corporation Superpixel edges for boundary detection
CN107067039A (en) * 2017-04-25 2017-08-18 西安电子科技大学 SAR image Ship Target quick determination method based on super-pixel
CN108629783A (en) * 2018-05-02 2018-10-09 山东师范大学 Image partition method, system and medium based on the search of characteristics of image density peaks
CN108765491A (en) * 2018-05-31 2018-11-06 成都信息工程大学 A kind of SAR image Ship Target Detection method
CN109886218A (en) * 2019-02-26 2019-06-14 西安电子科技大学 SAR image Ship Target Detection method based on super-pixel statistics diversity

Also Published As

Publication number Publication date
CN112766287B (en) 2021-09-17

Similar Documents

Publication Publication Date Title
CN110210463B (en) Precise ROI-fast R-CNN-based radar target image detection method
CN110378308B (en) Improved port SAR image near-shore ship detection method based on fast R-CNN
CN109871902B (en) SAR small sample identification method based on super-resolution countermeasure generation cascade network
CN109740460B (en) Optical remote sensing image ship detection method based on depth residual error dense network
CN108898065B (en) Deep network ship target detection method with candidate area rapid screening and scale self-adaption
CN106170819A (en) A kind of diameter radar image Ship Target method for quick
CN112418108B (en) Remote sensing image multi-class target detection method based on sample reweighing
CN113536963B (en) SAR image airplane target detection method based on lightweight YOLO network
CN114067118B (en) Processing method of aerial photogrammetry data
CN112508963B (en) SAR image segmentation method based on fuzzy C-means clustering
CN112651951A (en) DCE-MRI-based breast cancer classification method
CN115019187A (en) Detection method, device, equipment and medium for SAR image ship target
CN110930425B (en) Damaged target detection method based on neighborhood vector inner product local contrast image enhancement
CN112766286B (en) Ship target detection method based on global-local features of SAR image
Zhang et al. Contextual squeeze-and-excitation mask r-cnn for sar ship instance segmentation
CN114373135A (en) Ship target detection method based on local significance characteristic measurement
CN112766287B (en) SAR image ship target detection acceleration method based on density examination
CN117372676A (en) Sparse SAR ship target detection method and device based on attention feature fusion
CN111950476A (en) Deep learning-based automatic river channel ship identification method in complex environment
Li et al. A min-pooling detection method for ship targets in noisy SAR images
CN115205683A (en) Infrared small target detection method
CN115861669A (en) Infrared dim target detection method based on clustering idea
Wenxiu et al. Remote sensing ship detection technology based on DoG preprocessing and shape features
CN115187856A (en) SAR image ship detection method based on human eye vision attention mechanism
CN114998760A (en) Radar image ship detection network model and detection method based on domain adaptation

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