CN101900810A - Method for fusing multi-probe end sonar information by using submersible as carrier - Google Patents
Method for fusing multi-probe end sonar information by using submersible as carrier Download PDFInfo
- Publication number
- CN101900810A CN101900810A CN 201010226510 CN201010226510A CN101900810A CN 101900810 A CN101900810 A CN 101900810A CN 201010226510 CN201010226510 CN 201010226510 CN 201010226510 A CN201010226510 A CN 201010226510A CN 101900810 A CN101900810 A CN 101900810A
- Authority
- CN
- China
- Prior art keywords
- information
- probe
- probability
- sorter
- fusion
- 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
Links
Images
Abstract
The invention provides a method for fusing multi-probe end sonar information by using a submersible as a carrier. After received by a receiving end, echo information of a plurality of submersible probe ends of multi-probe end sonar equipment using the submersible as the carrier is fused according to the following steps of: (1) pretreating the echo information of the probe ends; (2) extracting various kinds of echo characteristic information of the probe ends; (3) fusing characteristic layers of the characteristic information; and (4) fusing decision-making layers of the fused information of the characteristic layers. In the information fusing method used in multi-probe end sonar using the submersible as the carrier, the information of the probe ends is systematically used for making decisions; the information of each probe end is fused; the characteristics of an underwater target are described comprehensively, accurately and reliably; the defects that a single probe end only provides the local characteristic information of the target and the characteristic of the target cannot be described comprehensively are overcome; and the accuracy and the reliability of the multi-probe end sonar equipment are enhanced.
Description
Technical field
What the present invention relates to is a kind of ocean sonar detection information processing method, particularly relates to a kind of information fusion method of many end of probe sonar.
Background technology
With small-sized latent device is that many end of probe of carrier sonar is a trend of marine survey technology equipment development, many end of probe sonar is made up of a plurality of underwater hiding-machine end (being end of probe) and receiving ends that carry sonar, and the echo information of system's end of probe is handled by unified reception of receiving end.Because the extremely complicacy of underwater environment in the Target Recognition process, is subjected to marine environment, reverberation and The noise bigger, to such an extent as to the accuracy and the reliability when relying on single information source to carry out Target Recognition of sonar detection are not high.Adopt a plurality of end of probe that target is surveyed, the echo of each end of probe is carried out information fusion.Utilize the echo information of each end of probe to greatest extent, remedy shortcomings such as the Target Recognition reliability that single information source causes is not high, accuracy is low.With obtain echo by the single detective end and carry out feature extraction and compare with Target Recognition, the echo information of many end of probe merges forme fruste, the uncertainty that can reduce the clarification of objective perception, improves the accuracy rate to Target Recognition.
Through existing technical literature retrieval is found, do not find identical with theme of the present invention or similar bibliographical information.
Summary of the invention
The object of the present invention is to provide a kind of feature that can describe submarine target comprehensively, accurately, reliably, improve the accuracy of many end of probe sonar and the multi-probe end sonar information by using submersible as carrier fusion method of reliability.
The object of the present invention is achieved like this:
The echo data information of a plurality of latent device end of probe that with the device of diving is many end of probe sonar of carrier is carried out fusion treatment according to following steps after receiving end receives:
(1) end of probe echo information data pre-service;
(2) end of probe echo various features information extraction;
(3) various features information characteristics layer merges;
(4) characteristic layer merges decision-making level's fusion of information.
The pretreated method of described end of probe echo information data is the wavelet transformation threshold method.
The information extraction of described end of probe echo various features comprises geometrical highlight feature extraction, the feature extraction of elasticity bright spot, position of orientation feature extraction; The geometrical highlight feature extracting method adopts sub belt energy feature extraction method; Elasticity bright spot feature extracting method adopts the discrete small wave converting method of signal spectrum; The position of orientation feature extracting method adopts multi-beam formation method.
Described various features information characteristics layer merges employing BP neural network design feature layer integrated classification device, the number of characteristic layer information fusion sorter is identical with the number of end of probe, the geometrical highlight feature, elasticity bright spot feature, the position of orientation feature that are input as extraction in the step (2) of each characteristic layer information fusion sorter; Be output as targeted species and orientation that corresponding end of probe detects.
The fusion of described decision-making level is a kind of high-level fusion to a plurality of sorter classification results, adopt probability weight voting scheme that the recognition result of a plurality of sorters is put to the vote, calculate each sorter to the possible probability of target sample identification, the possible probability that each sorter target azimuth sample is identical, the probability that compares each sorter classification results then, with the sample class of classification under the sample of probability maximum as correct identification, the big more explanation target of the probability that the target azimuth sample is identical is same target, otherwise is a plurality of targets.
Of the present invention is the information fusion method that uses in many end of probe sonar of carrier with the device of diving, the data message that utilizes a plurality of end of probe of system is made a strategic decision, the information data of each end of probe is merged, the feature of submarine target is described comprehensively, accurately, reliably, having overcome single end of probe can only provide the characteristic information of the part of target, can not comprehensively describe clarification of objective, improve the accuracy and the reliability of many end of probe sonar.
Characteristics of the present invention are mainly reflected in: the echo data information of a plurality of latent device end of probe of many end of probe sonar that with the device of diving is carrier is after receiving end receives, the present invention at first carries out the data pre-service to the echo information of each end of probe, remove interference and noise in the echo information, improve signal to noise ratio (S/N ratio).At its various features of information extraction of target, target is described then, the feature of extracting is sent into the integrated classification device of characteristic layer, carry out integrated classification at characteristic layer earlier, be output as the kind and the orientation of target from a plurality of angles.With the output of characteristic layer input, carry out integrated classification more then, be output as the number of target and the orientation of target correspondence in decision-making level as decision-making level.
The advantage of the inventive method is that by utilizing with the device of diving be the echo data information of a plurality of end of probe of carrier, the echo data information of each end of probe is carried out various features information extractions such as geometrical highlight, elasticity bright spot, position of orientation, carry out twice fusion at characteristic layer and decision-making level, utilize the echo information of each end of probe to greatest extent, remedy shortcomings such as the Target Recognition reliability that single information source causes is not high, accuracy is low.With obtain echo by the single detective end and carry out feature extraction and compare with Target Recognition, the echo information of many end of probe merges forme fruste, the uncertainty that can reduce the clarification of objective perception, improves the accuracy rate to Target Recognition.
Description of drawings
Fig. 1 is to be many end of probe sonar structural representation of carrier with the device of diving in the inventive method;
Fig. 2 is the schematic flow sheet of the inventive method.
Embodiment
For example the present invention is done description in more detail below in conjunction with accompanying drawing:
In conjunction with Fig. 1, Fig. 1 embodied is to be the structural representation of many end of probe sonar of carrier with the device of diving in the inventive method, having a plurality of is the end of probe 1 of carrier with the device of diving ... N (N determines according to the mission requirements of reality), the echo information 1 of end of probe ... N is received by unified processing of receiving end.
In conjunction with Fig. 2, information fusion method of the present invention is made up of following steps:
101, data message pretreatment stage: the data message S1 that receiving end is received N the end of probe of handling ... SN adopts the wavelet transformation threshold method to carry out the data pre-service, eliminates various interference of noise, improves signal to noise ratio (S/N ratio), the data I 1 after obtaining handling ... IN;
102, the data message various features is extracted the stage: to the pretreated data I 1 of step 101 ... IN carries out feature extraction, the feature of extraction comprises 11,12,13,21,22,23 ..., N1, N2, N3.11,12,13 position of orientation characteristic information, elasticity bright spot feature, the geometrical highlight features of representing respectively at the data message extraction of end of probe 1.21,22,23 position of orientation characteristic information, elasticity bright spot feature, the geometrical highlight features of representing respectively at the data message extraction of end of probe 2.N1, N2, N3 represent position of orientation characteristic information, elasticity bright spot feature, the geometrical highlight feature at the data message extraction of end of probe N respectively.The target of each end of probe being surveyed from a plurality of feature angles is described, and wherein the geometrical highlight feature extracting method adopts sub belt energy feature extraction method; Elasticity bright spot feature extracting method adopts the discrete small wave converting method of signal spectrum; The position of orientation feature extracting method adopts multi-beam formation method;
103, various features information characteristics layer fusing stage: with the characteristic information 11,12,13,21,22,23 that in the step 102 each end of probe extracted ..., N1, N2, N3 send into and adopt in the BP neural network design feature layer integrated classification device, obtains the kind and the azimuth information R1 of each end of probe institute detection of a target ... RN.The number of characteristic layer information fusion sorter is identical with the number of end of probe.With characteristic layer information fusion sorter 1 is example, it is input as the various features of in the step 102 end of probe 1 pretreated data message I1 being extracted 11,12 and 13, through obtaining the kind and the azimuth information R1 of end of probe 1 target that detects behind the characteristic layer information fusion sorter 1.
104, characteristic layer merges decision-making level's fusing stage of information, with the kind and the azimuth information R1 of resulting each end of probe institute detection of a target in the step 103 ... RN sends in the sorter that adopts probability weight voting conceptual design a plurality of characteristic layer integrated classification devices 1 ... the recognition result R1 of N ... RN puts to the vote.To a plurality of characteristic layer integrated classification devices 1 ... the classification results of N carries out high-level fusion, calculate each sorter to the possible probability of target sample identification, the possible probability that each sorter target azimuth sample is identical, the probability that compares each sorter classification results then, with the sample class of classification under the sample of probability maximum as correct identification, the big more explanation target of the probability that the target azimuth sample is identical is same target, otherwise is a plurality of targets.
Claims (9)
1. multi-probe end sonar information by using submersible as carrier fusion method is characterized in that with the device of diving being that the echo data information of a plurality of latent device end of probe of many end of probe sonar of carrier is carried out fusion treatment according to following steps after receiving end receives:
(1) end of probe echo information data pre-service;
(2) end of probe echo various features information extraction;
(3) various features information characteristics layer merges;
(4) characteristic layer merges decision-making level's fusion of information.
2. multi-probe end sonar information by using submersible as carrier fusion method according to claim 1 is characterized in that: the pretreated method of described end of probe echo information data is the wavelet transformation threshold method.
3. multi-probe end sonar information by using submersible as carrier fusion method according to claim 1 and 2 is characterized in that: the information extraction of described end of probe echo various features comprises geometrical highlight feature extraction, the feature extraction of elasticity bright spot, position of orientation feature extraction; The geometrical highlight feature extracting method adopts sub belt energy feature extraction method; Elasticity bright spot feature extracting method adopts the discrete small wave converting method of signal spectrum; The position of orientation feature extracting method adopts multi-beam formation method.
4. multi-probe end sonar information by using submersible as carrier fusion method according to claim 1 and 2, it is characterized in that: described various features information characteristics layer merges employing BP neural network design feature layer integrated classification device, the number of characteristic layer information fusion sorter is identical with the number of end of probe, the geometrical highlight feature, elasticity bright spot feature, the position of orientation feature that are input as extraction in the step (2) of each characteristic layer information fusion sorter; Be output as targeted species and orientation that corresponding end of probe detects.
5. multi-probe end sonar information by using submersible as carrier fusion method according to claim 3, it is characterized in that: described various features information characteristics layer merges employing BP neural network design feature layer integrated classification device, the number of characteristic layer information fusion sorter is identical with the number of end of probe, the geometrical highlight feature, elasticity bright spot feature, the position of orientation feature that are input as extraction in the step (2) of each characteristic layer information fusion sorter; Be output as targeted species and orientation that corresponding end of probe detects.
6. multi-probe end sonar information by using submersible as carrier fusion method according to claim 1 and 2, it is characterized in that: the fusion of described decision-making level is a kind of high-level fusion to a plurality of sorter classification results, adopt probability weight voting scheme that the recognition result of a plurality of sorters is put to the vote, calculate the possible probability of each sorter to target sample identification, the possible probability that each sorter target azimuth sample is identical, the probability that compares each sorter classification results then, with the sample class of classification under the sample of probability maximum as correct identification, the big more explanation target of the probability that the target azimuth sample is identical is same target, otherwise is a plurality of targets.
7. multi-probe end sonar information by using submersible as carrier fusion method according to claim 3, it is characterized in that: the fusion of described decision-making level is a kind of high-level fusion to a plurality of sorter classification results, adopt probability weight voting scheme that the recognition result of a plurality of sorters is put to the vote, calculate the possible probability of each sorter to target sample identification, the possible probability that each sorter target azimuth sample is identical, the probability that compares each sorter classification results then, with the sample class of classification under the sample of probability maximum as correct identification, the big more explanation target of the probability that the target azimuth sample is identical is same target, otherwise is a plurality of targets.
8. multi-probe end sonar information by using submersible as carrier fusion method according to claim 4, it is characterized in that: the fusion of described decision-making level is a kind of high-level fusion to a plurality of sorter classification results, adopt probability weight voting scheme that the recognition result of a plurality of sorters is put to the vote, calculate the possible probability of each sorter to target sample identification, the possible probability that each sorter target azimuth sample is identical, the probability that compares each sorter classification results then, with the sample class of classification under the sample of probability maximum as correct identification, the big more explanation target of the probability that the target azimuth sample is identical is same target, otherwise is a plurality of targets.
9. multi-probe end sonar information by using submersible as carrier fusion method according to claim 5, it is characterized in that: the fusion of described decision-making level is a kind of high-level fusion to a plurality of sorter classification results, adopt probability weight voting scheme that the recognition result of a plurality of sorters is put to the vote, calculate the possible probability of each sorter to target sample identification, the possible probability that each sorter target azimuth sample is identical, the probability that compares each sorter classification results then, with the sample class of classification under the sample of probability maximum as correct identification, the big more explanation target of the probability that the target azimuth sample is identical is same target, otherwise is a plurality of targets.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201010226510 CN101900810A (en) | 2010-07-15 | 2010-07-15 | Method for fusing multi-probe end sonar information by using submersible as carrier |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201010226510 CN101900810A (en) | 2010-07-15 | 2010-07-15 | Method for fusing multi-probe end sonar information by using submersible as carrier |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101900810A true CN101900810A (en) | 2010-12-01 |
Family
ID=43226503
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201010226510 Pending CN101900810A (en) | 2010-07-15 | 2010-07-15 | Method for fusing multi-probe end sonar information by using submersible as carrier |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101900810A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102158551A (en) * | 2011-03-30 | 2011-08-17 | 沈益民 | Information source unified description and access method for Internet of things |
CN102692625A (en) * | 2012-05-15 | 2012-09-26 | 哈尔滨工程大学 | Joint modeling method for features of underwater target echo and reverberation in Rn space |
CN103901422A (en) * | 2014-03-21 | 2014-07-02 | 哈尔滨工程大学 | Underwater target echo geometric bright spot structure characteristic extracting method |
CN104598624A (en) * | 2015-02-04 | 2015-05-06 | 苏州大学 | User class determination method and device for microblog user |
CN107590468A (en) * | 2017-09-15 | 2018-01-16 | 哈尔滨工程大学 | A kind of detection method based on various visual angles target highlight feature fusion |
CN110441761A (en) * | 2019-09-18 | 2019-11-12 | 哈尔滨工程大学 | Multi-sources Information Fusion Method based on the detection of distributed buoy |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5828625A (en) * | 1997-10-09 | 1998-10-27 | The United States Of America As Represented By The Secretary Of The Navy | Echo simulator for active sonar |
CN101149436A (en) * | 2007-11-06 | 2008-03-26 | 哈尔滨工程大学 | Super broad coverage multiple beam bathymetric side scanning sonar device |
JP4288663B2 (en) * | 2003-06-05 | 2009-07-01 | 日本電気株式会社 | Phase phasing method and apparatus |
CN101650436A (en) * | 2009-09-23 | 2010-02-17 | 哈尔滨工程大学 | Embedded type intelligent acoustic detection system |
US7719461B1 (en) * | 2008-08-05 | 2010-05-18 | Lockheed Martin Corporation | Track fusion by optimal reduced state estimation in multi-sensor environment with limited-bandwidth communication path |
-
2010
- 2010-07-15 CN CN 201010226510 patent/CN101900810A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5828625A (en) * | 1997-10-09 | 1998-10-27 | The United States Of America As Represented By The Secretary Of The Navy | Echo simulator for active sonar |
JP4288663B2 (en) * | 2003-06-05 | 2009-07-01 | 日本電気株式会社 | Phase phasing method and apparatus |
CN101149436A (en) * | 2007-11-06 | 2008-03-26 | 哈尔滨工程大学 | Super broad coverage multiple beam bathymetric side scanning sonar device |
US7719461B1 (en) * | 2008-08-05 | 2010-05-18 | Lockheed Martin Corporation | Track fusion by optimal reduced state estimation in multi-sensor environment with limited-bandwidth communication path |
CN101650436A (en) * | 2009-09-23 | 2010-02-17 | 哈尔滨工程大学 | Embedded type intelligent acoustic detection system |
Non-Patent Citations (2)
Title |
---|
《中国优秀硕士学位论文全文数据库工程科技II辑》 20071015 杨秀雷 自主推进式猎雷声纳系统体系结构研究 第13,15,36页,图4.4 1-9 , * |
《声学技术》 20050930 郭丽华等 水下目标特征提取方法研究 第24卷, 第3期 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102158551A (en) * | 2011-03-30 | 2011-08-17 | 沈益民 | Information source unified description and access method for Internet of things |
CN102692625A (en) * | 2012-05-15 | 2012-09-26 | 哈尔滨工程大学 | Joint modeling method for features of underwater target echo and reverberation in Rn space |
CN103901422A (en) * | 2014-03-21 | 2014-07-02 | 哈尔滨工程大学 | Underwater target echo geometric bright spot structure characteristic extracting method |
CN104598624A (en) * | 2015-02-04 | 2015-05-06 | 苏州大学 | User class determination method and device for microblog user |
CN107590468A (en) * | 2017-09-15 | 2018-01-16 | 哈尔滨工程大学 | A kind of detection method based on various visual angles target highlight feature fusion |
CN107590468B (en) * | 2017-09-15 | 2020-07-24 | 哈尔滨工程大学 | Detection method based on multi-view target bright spot characteristic information fusion |
CN110441761A (en) * | 2019-09-18 | 2019-11-12 | 哈尔滨工程大学 | Multi-sources Information Fusion Method based on the detection of distributed buoy |
CN110441761B (en) * | 2019-09-18 | 2023-04-11 | 哈尔滨工程大学 | Multi-source information fusion method based on distributed buoy detection |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101900810A (en) | Method for fusing multi-probe end sonar information by using submersible as carrier | |
US9467317B2 (en) | Method and apparatus for detecting and classifying signals | |
CN102279390B (en) | Intra-pulse modulation and recognition method of low signal-to-noise radar radiation source signal | |
CN101470194B (en) | Torpedo target recognition method | |
Martin et al. | Estimating minke whale (Balaenoptera acutorostrata) boing sound density using passive acoustic sensors | |
Borhani-Darian et al. | Deep neural network approach to detect GNSS spoofing attacks | |
CN106483514B (en) | Airplane motion mode identification method based on EEMD and support vector machine | |
CN103093244B (en) | A kind of Radar Signal Recognition method based on Its Sparse Decomposition | |
CN111985349B (en) | Classification recognition method and system for radar received signal types | |
Murphy et al. | Examining the robustness of automated aural classification of active sonar echoes | |
CN107817484A (en) | The multiplication factor processing method and processing device of laser radar amplifying circuit | |
CN114155445A (en) | SAR image target detection method based on improved YOLOv3 | |
JP7143809B2 (en) | Clutter learning device and clutter discriminating device | |
CN113484838B (en) | Multi-base radar active false target identification method and system | |
CN117233706B (en) | Radar active interference identification method based on multilayer channel attention mechanism | |
CN106199598A (en) | A kind of maritime peril target homing system and method based on marine radar | |
CN102540159A (en) | Method for detecting small targets of sea surface by utilizing least squares support vector machine (LS-SVM) on basis of wavelet noise reduction | |
CN115932895A (en) | GNSS deception jamming intelligent detection method based on MHA-1DCNN | |
JP6180065B2 (en) | Object position detection apparatus and method | |
CN113759362B (en) | Method, device, equipment and storage medium for radar target data association | |
CN103308918B (en) | Fish identification method and system based on segmented time-domain centroid features | |
CN102122453B (en) | Simulation system and method of electronic interference nacelle | |
Rahman et al. | Ensemble classifier based modulation recognition for beyond 5G massive MIMO (mMIMO) communication | |
CN113376586A (en) | Method for constructing classification model of double-component radar signals | |
CN112666528A (en) | Multi-station radar system interference identification method based on convolutional neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20101201 |