CN105930847A - Combined edge detection-base SAR image linear feature extraction method - Google Patents

Combined edge detection-base SAR image linear feature extraction method Download PDF

Info

Publication number
CN105930847A
CN105930847A CN201610211874.0A CN201610211874A CN105930847A CN 105930847 A CN105930847 A CN 105930847A CN 201610211874 A CN201610211874 A CN 201610211874A CN 105930847 A CN105930847 A CN 105930847A
Authority
CN
China
Prior art keywords
edge detection
sar image
linear
edge
adopted
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.)
Pending
Application number
CN201610211874.0A
Other languages
Chinese (zh)
Inventor
杨桄
程红
谭海峰
杨永波
王志强
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.)
AIR FORCE AVIATION UNIVERSITY OF CHINESE PEOPLE'S LIBERATION ARMY
Original Assignee
AIR FORCE AVIATION UNIVERSITY OF CHINESE PEOPLE'S LIBERATION ARMY
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 AIR FORCE AVIATION UNIVERSITY OF CHINESE PEOPLE'S LIBERATION ARMY filed Critical AIR FORCE AVIATION UNIVERSITY OF CHINESE PEOPLE'S LIBERATION ARMY
Priority to CN201610211874.0A priority Critical patent/CN105930847A/en
Publication of CN105930847A publication Critical patent/CN105930847A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/40Extraction of image or video features
    • G06V10/44Local 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
    • G06V10/457Local 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 by analysing connectivity, e.g. edge linking, connected component analysis or slices

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a combined edge detection-base SAR image linear feature extraction method. The method comprises the following steps that: a Frost algorithm is adopted to perform filtering preprocessing on an image; a compromise is made between speckle noise removal and edge information reservation; a Canny operator and ROA operator-based edge detection algorithm is adopted to carry out image edge detection, wherein the Canny operator and ROA operator-based edge detection algorithm has a great edge positioning ability when maintaining a constant false alarm rate; an improved Radon method is adopted to carry out linear support set extraction, the least square method is adopted to fit a linear support set so as to obtain linear primitives; and fractured linear primitives are connected together based on a point heuristic type connection thought, and extracted linear features can be further improved. With the combined edge detection-base SAR image linear feature extraction method of the invention adopted, man-machine interaction, accuracy and convenience of SAR image detection can be detected.

Description

SAR image linear characteristic extracting method based on combination of edge detection
Technical field
The present invention relates to image co-registration field, be specifically related to a kind of SAR image line based on combination of edge detection Property feature extracting method.
Background technology
SAR image linear characteristic extracting method generally comprises 3 steps: be first filtered SAR image Deng pretreatment;Then neighborhood of pixels is detected and marginal point is separated with background area, referred to as rim detection; Finally the marginal point in units of pixel is connected into straight line and describes its line structure, referred to as edge marshalling.Limit Edge detection is Pixel-level computing.In the rim detection of optical imagery, generally use Canny or zero crossing etc. Gradient operator.Wherein, Canny operator has preferable edge stationkeeping ability, and operand is moderate, but for It it is Additive noise model.And the noise of SAR image is the multiplicative noise that approximation obeys Gamma distribution, therefore, Difference gradient operator is no longer CFAR for SAR image.Touzi et al. proposes, ratio of averages (Ratio Of Average, ROA) operator has the characteristic of CFAR in the coherent spot that Gamma is distributed, But ROA operator edge stationkeeping ability is poor, it is impossible to the edge inspection being directly used in during linear character automatically extracts Survey.Edge marshalling is generally divided into two steps: line primitives extracts and line primitives connects.Line primitives extract typically have with Lower several method: the extraction of phase coding sequences, line primitives based on template, Radon conversion etc..From line primitives From the point of view of extraction ratio, correctness, extraction quality etc., the extraction performance of Radon conversion is best, and Radon Convert the directional information without using marginal point, overcome low resolution SAR image edge direction location inaccurate Problem.Often there is defect and deviation in the rectilinear after line primitives extracts, needs the line base of fracture for this Unit couples together.
Summary of the invention
For solving the problems referred to above, the invention provides a kind of SAR image based on combination of edge detection the most special Levy extracting method, add man-machine interaction, accuracy and convenience during SAR image detection.
For achieving the above object, the technical scheme that the present invention takes is:
SAR image linear characteristic extracting method based on combination of edge detection, comprises the steps:
S1, employing Frost algorithm are filtered pretreatment to image;Remove speckle noise and keep edge Traded off between information.
S2, employing edge detection algorithm based on Canny operator and ROA operator carry out Image Edge-Detection; This algorithm has preferable edge stationkeeping ability while keeping CFAR.
The Radon method that S3, employing improve carries out straight line support collection and extracts and use least square method to directly Line support collection is fitted, and obtains line primitives;
S4, the line primitives of fracture is coupled together based on putting heuristic connection thought, the most perfect extract Linear feature.It is demonstrated experimentally that new lines detection method preferably describes the linear spy in SAR image Levy.
Wherein, described step S4 specifically includes following steps: first design is suitable for the adaptation of Radon conversion Degree function, then regards line primitives marginal point (i.e. node) as, the extreme coordinates of line primitives, slope etc. is believed Cease the state as node, utilize fitness function to be coupled together by the node with least disadvantage path.
The method have the advantages that
For the feature of low signal-to-noise ratio SAR image, have chosen a kind of new linear character being suitable to Project Realization Extracting method.Canny operator and the edge detection algorithm of ROA operator is merged, it is achieved that detection by using CFAR also overcomes line interruption, and detection algorithm has higher edge stationkeeping ability.Radon is utilized to convert Extract line primitives, it is not necessary to CONSIDERING EDGE direction, overcome edge direction location that low signal-to-noise ratio brings inaccurate Problem.Heuristic connection thought based on point, defines fitness function based on Radon conversion, it is achieved Being reliably connected of line primitives.By the form of software, method is showed, user-friendly, soft Part operation result shows: the linear character accurate description of the extraction linear structure of SAR image, the line obtained Property feature can be used for the aspect such as automatic target detection and scene matching aided navigation.
Accompanying drawing explanation
Fig. 1 is the stream of the SAR image linear characteristic extracting method that the embodiment of the present invention detects based on combination of edge Cheng Tu.
The SAR image linear characteristic extracting method that Fig. 2 is the embodiment of the present invention to be detected based on combination of edge walks The flow chart of rapid S4.
Detailed description of the invention
In order to make objects and advantages of the present invention clearer, below in conjunction with embodiment, the present invention is carried out Further describe.Should be appreciated that specific embodiment described herein only in order to explain the present invention, It is not intended to limit the present invention.
As it is shown in figure 1, embodiments provide SAR image linear character based on combination of edge detection Extracting method, comprises the steps:
S1, employing Frost algorithm are filtered pretreatment to image;Remove speckle noise and keep edge Traded off between information.
S2, employing edge detection algorithm based on Canny operator and ROA operator carry out Image Edge-Detection; This algorithm has preferable edge stationkeeping ability while keeping CFAR.
The Radon method that S3, employing improve carries out straight line support collection and extracts and use least square method to directly Line support collection is fitted, and obtains line primitives;
S4, the line primitives of fracture is coupled together based on putting heuristic connection thought, the most perfect extract Linear feature.It is demonstrated experimentally that new lines detection method preferably describes the linear spy in SAR image Levy.
Owing to SAR image existing coherent speckle noise, long line interruption is usually made to become several sections.In order to obtain Complete linear character, needs to couple together the straightway of fracture.This process is referred to as line primitives and connects. Line primitives connection establishment is on the basis of extracting more complete line primitives.
SAR image signal to noise ratio is low, coherent spot strong, typically can not draw the accurate direction of marginal point.The present invention The algorithm with certain geometry analyticity is used marginal point to be gathered into straight line support collection, then to straight line support Collection is fitted realizing line primitives and extracts.Radon conversion has clear geometry analyticity, certain resisting is done Disturb ability and be prone to the advantages such as parallel processing, and, compared with Hough transform, Radon transformation calculations amount Little, need not the biggest memory space.So, the present invention uses Radon conversion to realize straight line support collection and extracts.
Heuristic connection is usually the Edge track of figure, and edge graph Shang You edge is all node, the width at edge Degree and phase place are the state of node, and attended operation is the set of node having least disadvantage path in search graph.Minimum The definition of loss is relevant with fitness function used, also relevant with concrete task.Not only to consider existing The relation of node and next both candidate nodes, it is also contemplated that the situation of the upper node that it is connected with existing node.
The present invention uses the thought of heuristic connection to be coupled together by the line primitives of fracture, as in figure 2 it is shown, tool Body comprises the steps: first to design the fitness function of applicable Radon conversion, is then regarded as by line primitives Marginal point (i.e. node), using information such as the extreme coordinates of line primitives, slopes as the state of node, utilizes suitable The node with least disadvantage path is coupled together by response function.
The above is only the preferred embodiment of the present invention, it is noted that common for the art For technical staff, under the premise without departing from the principles of the invention, it is also possible to make some improvements and modifications, These improvements and modifications also should be regarded as protection scope of the present invention.

Claims (2)

1. SAR image linear characteristic extracting method based on combination of edge detection, it is characterised in that include Following steps:
S1, employing Frost algorithm are filtered pretreatment to image;
S2, employing edge detection algorithm based on Canny operator and ROA operator carry out Image Edge-Detection;
The Radon method that S3, employing improve carries out straight line support collection and extracts and use least square method to directly Line support collection is fitted, and obtains line primitives;
S4, the line primitives of fracture is coupled together based on putting heuristic connection thought, the most perfect extract Linear feature.
SAR image Linear feature extraction side based on combination of edge detection the most according to claim 1 Method, it is characterised in that described step S4 specifically includes following steps: first design is suitable for Radon conversion Fitness function, then regard line primitives as marginal point, by information such as the extreme coordinates of line primitives, slopes As the state of node, fitness function is utilized to be coupled together by the node with least disadvantage path.
CN201610211874.0A 2016-03-31 2016-03-31 Combined edge detection-base SAR image linear feature extraction method Pending CN105930847A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610211874.0A CN105930847A (en) 2016-03-31 2016-03-31 Combined edge detection-base SAR image linear feature extraction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610211874.0A CN105930847A (en) 2016-03-31 2016-03-31 Combined edge detection-base SAR image linear feature extraction method

Publications (1)

Publication Number Publication Date
CN105930847A true CN105930847A (en) 2016-09-07

Family

ID=56840188

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610211874.0A Pending CN105930847A (en) 2016-03-31 2016-03-31 Combined edge detection-base SAR image linear feature extraction method

Country Status (1)

Country Link
CN (1) CN105930847A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107016397A (en) * 2017-04-25 2017-08-04 深圳市大德激光技术有限公司 Local template intelligently inspires global image matching process
CN107886505A (en) * 2017-11-08 2018-04-06 电子科技大学 A kind of synthetic aperture radar airfield detection method based on line primitives polymerization
CN110782471A (en) * 2019-10-16 2020-02-11 中国矿业大学 Multi-scale SAR image edge detection method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102521802A (en) * 2011-11-28 2012-06-27 广东省科学院自动化工程研制中心 Mathematical morphology and LoG operator combined edge detection algorithm
CN104200442A (en) * 2014-09-19 2014-12-10 西安电子科技大学 Improved canny edge detection based non-local means MRI (magnetic resonance image) denoising method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102521802A (en) * 2011-11-28 2012-06-27 广东省科学院自动化工程研制中心 Mathematical morphology and LoG operator combined edge detection algorithm
CN104200442A (en) * 2014-09-19 2014-12-10 西安电子科技大学 Improved canny edge detection based non-local means MRI (magnetic resonance image) denoising method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
安成锦等: "基于融合边缘检测的SAR图像线性特征提取算法", 《电子与信息学报》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107016397A (en) * 2017-04-25 2017-08-04 深圳市大德激光技术有限公司 Local template intelligently inspires global image matching process
CN107886505A (en) * 2017-11-08 2018-04-06 电子科技大学 A kind of synthetic aperture radar airfield detection method based on line primitives polymerization
CN110782471A (en) * 2019-10-16 2020-02-11 中国矿业大学 Multi-scale SAR image edge detection method

Similar Documents

Publication Publication Date Title
CN103927526B (en) Vehicle detecting method based on Gauss difference multi-scale edge fusion
Wu et al. Lane-mark extraction for automobiles under complex conditions
KR101609303B1 (en) Method to calibrate camera and apparatus therefor
CN103500322B (en) Automatic lane line identification method based on low latitude Aerial Images
Daigavane et al. Road lane detection with improved canny edges using ant colony optimization
CN102982304B (en) Utilize polarized light image to detect the method and system of vehicle location
CN105426864A (en) Multiple lane line detecting method based on isometric peripheral point matching
CN102999886B (en) Image Edge Detector and scale grating grid precision detection system
CN103196372B (en) A kind of optical imagery detection method of electrification railway contact net supportive device
CN104008387A (en) Lane line detection method based on feature point piecewise linear fitting
CN106529493A (en) Robust multi-lane line detection method based on perspective drawing
CN101620732A (en) Visual detection method of road driving line
CN104239867A (en) License plate locating method and system
CN103383733A (en) Lane video detection method based on half-machine study
CN103544491A (en) Optical character recognition method and device facing complex background
CN105807335A (en) Vehicle chassis inspection method and system
CN101916373A (en) Road semiautomatic extraction method based on wavelet detection and ridge line tracking
Liu et al. ROI perspective transform based road marking detection and recognition
CN105930847A (en) Combined edge detection-base SAR image linear feature extraction method
CN104463138A (en) Text positioning method and system based on visual structure attribute
CN106327464A (en) Edge detection method
CN109447093A (en) Tail-light detection method based on YUV image
CN103886609A (en) Vehicle tracking method based on particle filtering and LBP features
CN103914829A (en) Method for detecting edge of noisy image
CN104732530A (en) Image edge detection method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20160907

RJ01 Rejection of invention patent application after publication