CN103985385A - Method for identifying Batrachia individual information based on spectral features - Google Patents
Method for identifying Batrachia individual information based on spectral features Download PDFInfo
- Publication number
- CN103985385A CN103985385A CN201410238578.0A CN201410238578A CN103985385A CN 103985385 A CN103985385 A CN 103985385A CN 201410238578 A CN201410238578 A CN 201410238578A CN 103985385 A CN103985385 A CN 103985385A
- Authority
- CN
- China
- Prior art keywords
- batrachia
- spectral characteristic
- feature database
- pipes
- record
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a method for identifying Batrachia individual information based on spectral features. According to the method, the Batrachia individual information can be distinguished and output according to Batrachia sound data collected in the outdoor environment, and the accurate reference data can be provided for richness research and individual trace research of Batrachia populations. The method comprises the following steps that firstly, a feature library is built, in the feature library, each tested Batrachia species corresponds to at least one record, each record comprises sample spectral feature values, and each spectral feature value represents the species, the gender, the health state, the courtship state and other features of the object corresponding to the spectral feature value; secondly, the tested data are processed; thirdly, comparison and judgment are performed on the tested data and the feature library.
Description
Technical field
The present invention relates to species authentication technique, be specifically related to a kind of batrachia individual information authentication technique.
Background technology
Occurring in nature, batrachia can send different piping by it and carry out acoustics communication, it pipes and is transmitting abundant information, comprise the information such as species identification, reproductive status, present position, attraction and selection opposite sex spouse and Individual Size, batrachia not of the same race pipes and has the singularity of its cry.By the identification to different batrachias, be conducive to identify different types of frog in people's lowered in field environment, be also the richness of observation batrachia population, pursue one of main clue of its trace, there is reality and actual demand.
At present, the analysis that pipes of batrachia is mainly to rely on artificial cognition.Subjective consciousness is stronger, still lacks suitable method and standard.
Summary of the invention
Technical matters solved by the invention is to provide a kind of method of identifying batrachia individual information based on Spectral Characteristic, the batrachia voice data that can collect according to field environment, identify and export the individual information of batrachia, can provide reference data accurately for the research of batrachia abundance of microbiological population, the research of individual trace.
For solving the problems of the technologies described above, the method for identifying batrachia individual information based on Spectral Characteristic of the present invention, comprises the following steps:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
How many according to the classification of tested frog kind, repeating step A is to step C, sets up the batrachia feature database that pipes; In described feature database, each tested frog kind is to there being at least one record; In every record, comprise sample Spectral Characteristic value, each Spectral Characteristic value has characterized its corresponding object kind;
Step 2: process measured data;
By the piping through noise reduction, cut apart and be extracted as measured data of needs identification, obtain the Spectral Characteristic value of measured data;
Step 3: determination step;
By HMM, calculate the similarity degree of measured data and feature database the inside species, export name title and the Spectral Characteristic value thereof of the most similar a kind of batrachia.
Preferably, also comprise:
Step 4: step 3 is augmented in feature database by the correct test data of real example.
Preferably, in described step 3, all records in measured data and feature database are compared, all records that are consistent with measured data in marker characteristic storehouse, and in all records that are labeled, the highest attribute of the frequency of occurrences is judged as the attribute being consistent with measured data.
Preferably, in the record of described feature database, each Spectral Characteristic value has also characterized its corresponding Individual Size feature, individual health status flag; In described step 3, first judge frog kind, then judge measurand Individual Size feature, individual health status flag.
Preferably, in the record of described feature database, each Spectral Characteristic value has also characterized its corresponding sex information; In described step 3, also comprise the step of judging measurand sex character.
Preferably, in the record of described feature database, each Spectral Characteristic value has also characterized its corresponding Reproductive State information; In described step 3, also comprise the step of judging measurand Reproductive State.
Preferably, in described step 3, measured data only with feature database in sex be that male record is compared.
Preferably, in the record of described feature database, the corresponding unique individuality numbering of each male, in described determination step, also comprises individual determination step.
Accompanying drawing explanation
Fig. 1 is the identifying process flow diagram of identifying the method for batrachia individual information based on Spectral Characteristic of the present invention;
Fig. 2 is the wave spectrogram of a complete pulse of Rana nigromaculata tweeting sound of the male breeding period of the larger individuality of a certain health in feature database;
Fig. 3 is the cepstrum figure of a complete pulse of Rana nigromaculata tweeting sound of the male breeding period of the larger individuality of a certain health in feature database;
Fig. 4 is the spectrogram of a complete pulse of Rana nigromaculata tweeting sound of the male breeding period of the larger individuality of a certain health in feature database;
Fig. 5 is the wave spectrogram of a complete pulse of bullfrog tweeting sound of the male breeding period of the larger individuality of a certain health in feature database;
Fig. 6 is the cepstrum figure of a complete pulse of bullfrog tweeting sound of the male breeding period of the larger individuality of a certain health in feature database;
Fig. 7 is the spectrogram of a complete pulse of bullfrog tweeting sound of the male breeding period of the larger individuality of a certain health in feature database.
Embodiment
Embodiment mono-
The kind of information of the affiliated object of tested sound be differentiated and be exported to the present embodiment can, according to the known voice data in the source collecting.
The present embodiment comprises the following steps:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
Repeating step A is to step C so repeatedly, sets up the batrachia feature database that pipes; In the record of described feature database, comprise sample Spectral Characteristic value, each Spectral Characteristic value has characterized its corresponding object kind.
Step 2: piping of needs identification done to identical processing to obtain Spectral Characteristic value; The data of obtaining are called measured data.
Step 3: calculate the pipe similarity degree of feature database the inside species of batrachia to be detected and batrachia by HMM, export name title and the Spectral Characteristic value thereof of the most similar a kind of batrachia.
Illustrate:
Below Rana nigromaculata, bullfrog, Rana temporaria chensinensis, rana limnocharis, piping of the brave line frog are identified.
First, gather the tweeting sound of different batrachias, in cry signal, at least comprise the tweeting sound of a complete frog.Gatherer process can carry out in the wild, under the situation satisfying the requirements, can carry out in laboratory.
Then, the sound that pipes of the frog is carried out to noise reduction process, in the sound that pipes after noise reduction, intercept a complete fragment of tweeting sound clearly.
Tweeting sound fragment is analyzed, extract MFCC feature, set up the property data base that pipes of batrachia, in described database, the characteristic that has at least comprised Rana nigromaculata, bullfrog, Rana temporaria chensinensis, rana limnocharis, the brave line frog, as visible in Fig. 2 to Fig. 7, characteristic has comprised the Spectral Characteristic of each frog kind; In property data base, each frog kind includes at least 1 record; In follow-up decision procedure, according to the size of the transition probability of sound in tested target voice and property data base, judge similarity.
The sound that pipes of test use, we select Rana nigromaculata (the sound feature that pipes of Rana nigromaculata has existed the property data base that pipes) to carry out the prediction accuracy of test macro.The wave spectrogram that described Spectral Characteristic value has comprised a complete pulse, cepstrum figure, spectrogram.As can be seen, the present invention has realized digitizing by the croaking of a frog of analog quantity, and those skilled in the art can realize by known algorithm comparing of recording in measured data and feature database on this basis; By the adjustment to parameter, can realize the adjustment of comparison accuracy.
We select 100 Rana nigromaculatas and bullfrog tweeting sound, the identification degree of test macro to Rana nigromaculata and bullfrog tweeting sound, through identification test, more than 90% Rana nigromaculata and bullfrog obtain the correct identification of system, illustrate that this method has a good application prospect, can reach each ergonomist to the pipe demand of feature and kind identification of batrachia.
In the field that the cry of animal is analyzed, both there is no a kind of general method for mode matching, there is no unified Spectral Characteristic extraction scheme yet.Particularly consider that amphibian has the aquatic and amphibious characteristic of Lu Sheng, its background noise that pipes is larger, about batrachia pipe feature automatic identification and identify, still there is not at present a kind of effective processing scheme.And, wave spectrum identification to other higher mammals (as cat and dog), it is all the peak value of LPCC or even frequency that the extraction of its feature that pipes is used, the more difficult level of section to sow that recognize of these features, because same section, wave spectrum difference between variety classes is relatively little, adds the stronger background noise of aquatic animal, makes its identification become more difficult.At present, there is not yet about analyzing the pipe method of individual information of batrachia, and on mode identification method, existing method adopts the algorithm of artificial intelligence network (ANN) and Non-negative Matrix Factorization (NMF) more, but degree of ripeness, reliability and robustness also need further confirmation.
The present invention has adopted hidden Markov model (Hidden Markov Model, HMM) to identify.Hidden Markov model (HMM) is a kind of probabilistic model based on transition probability and output probability, can embody the spectral characteristic and the time variation that pipe, can characterize preferably the Spectral Characteristic of batrachia.This model can be described the dynamic change of wave spectrum and the statistical distribution characteristic of wave spectrum preferably, the difference that species taxonomical unit section and following unit pipe between planting be can simulate efficiently, the relatively little species of wave spectrum difference pipe signature analysis and kind identification are applicable between batrachia not of the same race.The distinctive robustness of HMM simultaneously, also for the cry that gathers under the more environment of the interference such as field acquisition and identification and the analysis of variety classes (the following level of section) individual state provide effective guarantee.
As shown in Figure 1, the present invention has further adopted the vector quantization based on K mean cluster analysis algorithm at the front end of HMM, carries out the design of code book, then sets up hidden Marko Lip river husband's training pattern.Key step is to set up the descriptive data structure of HMM; Calculate one or more given observation sequence, i.e. the parameter such as the forward direction probability of each sound characteristic parameter, backward probability; HMM parameter initialization; Use Viterbi algorithm to train; Adopt Viterbi algorithm to identify; Viterbi algorithm is wherein a kind of algorithm more classical in voice recognition.
The pipe acquisition approach of characteristic of batrachia to be detected in the present invention comprises that terminal device does local analytics to piping, and extracts eigenwert, uploads onto the server and carries out similarity calculating; Or then file to the server of directly uploading batrachia tweeting sound by terminal device is processed.Used terminal can include but not limited to mobile phone, personal digital assistant, and wireless handheld equipment, gets online without being tethered to a cable this, PC, convenient computer, MP3 player, MP4 player etc.
This algorithm also can be integrated into related hardware device, and this device can gather tweeting sound signal, also can obtain tweeting sound file by network, USB interface etc.This device carries the batrachia sound feature database that pipes, can local analytics, calculate tweeting sound to be detected, export the tweeting sound (one or more) of most probable batrachia species.
In various embodiments of the present invention, all by the correct test data of real example can typing feature database in, for follow-up test provides foundation.That is to say, along with the expansion of feature database, what one species was corresponding records many, in carrying out the process of measurand differentiation, each measured data can with feature database in all records compare, and may simultaneously there is accordance with many data; In test process, all records that are consistent with measured data in marker characteristic storehouse, in all records that are labeled, certain the highest field attribute of the frequency of occurrences is judged as the respective attributes of measured data.For example: judge that in the process of frog kind, measured data conforms to some records in feature database simultaneously, wherein point to the maximum of A frog kind, judge that measured data meets the maximum probability of A frog kind.Feature database of the present invention, it records extendible, and along with by the increase of real example data, system accuracy will improve constantly.
Embodiment bis-
On the basis of embodiment mono-, the present embodiment further provides the information such as the Individual Size, health status of tested batrachia.
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
Described pattern batrachia comprises the various frogs of various states, such as the larger individuality of health, unsound larger individuality, healthy less individuality, unsound less individuality etc.; The frog of every kind of state has respectively several samples.
Described data can acquired original, also can carry out typing after real example according to the data of embodiment mono-.
How many according to the classification of tested frog kind, repeating step A is to step C, sets up the batrachia feature database that pipes; In described feature database, each tested frog kind is to there being at least one record; In every record, comprise sample Spectral Characteristic value, each Spectral Characteristic value has characterized its corresponding object species characteristic, has also characterized its corresponding Individual Size feature, individual health status flag;
Step 2: tweeting sound to be detected is done to identical noise reduction, cut apart, extract eigenwert;
Step 3: the similarity degree that calculates batrachia to be detected and batrachia tweeting sound feature database the inside species by HMM, according to detecting tweeting sound and the corresponding situation recording in feature database, export the title of the most similar batrachia.
Determining under the prerequisite of frog kind, according to measured data and the corresponding situation recording in feature database, the signs such as the health degree of output corresponding record institute mark and Individual Size.In this step, first determine frog kind, then incongruent frog kind record is got rid of, in remainder record, determine the attributes such as the corresponding individual size of measured data, health.
In the present embodiment, measured data is same typing feature database after real example is correct.Feature database comprises the indexs such as Spectral Characteristic value, frog kind, Individual Size, health status.
Embodiment tri-
On the basis of embodiment mono-, the present embodiment further provides the sex information of tested batrachia.
The present embodiment comprises the following steps:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
Repeating step A to step C, preserves various parameters, sets up the batrachia feature database that pipes; Each Spectral Characteristic value has characterized its corresponding object species characteristic, has also further characterized its corresponding sex character.
The batrachia sound of different sexes is not identical, such as Rana nigromaculata: male louder and clearer, female sound is more overcast.Above feature can be identified effectively by people's ear, adopts the MFCC method of listening force characteristic based on people's ear, can effectively capture these features.
Step 2: piping of needs identification done to identical processing to obtain Spectral Characteristic value.
Step 3: calculate the pipe similarity degree of feature database the inside species of batrachia to be detected and batrachia by HMM, export the name title of the most similar a kind of batrachia, and export the most similar sex title.
In the present embodiment, the same typing feature database of data that real example is correct.
Embodiment tetra-
On the basis of embodiment tri-, the present embodiment further provides measurand Reproductive State method of discrimination.
The present embodiment comprises:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
Repeating step A to step C, preserves various parameters, sets up the batrachia feature database that pipes; Each Spectral Characteristic value has characterized its corresponding object species characteristic, has also further characterized its corresponding sex character, with and Reproductive State feature.
In mating period, pipe relevant with reproduction from mating is different, for example, mainly comprises that with piping of the relevant Rana nigromaculata of seeking a spouse advertisement pipes, replys and pipe, refuse to pipe and acceptance pipes etc.It is mainly to be sent by male that advertisement pipes, attract female or declare its manor of occupying, announce the information such as self Individual Size to other rivals, and be the modal type that pipes.Replying pipes is that the male frog sends after advertisement pipes and obtains after the acceptance of the female frog, and the another kind then sending is sought a spouse and piped.The acceptance that is intended to answer the female frog pipes, and echoes with it, right to prepare to embrace.Acceptance pipes, and to be female Rana nigromaculata reply in breeding period a kind of the seeking a spouse that male frog advertisement pipes pipes, and represents to accept the wish of seeking a spouse of the male frog, pipes overcast.It is piping of sending while refusing the courtship ritual of the male frog of the female frog that refusal pipes, the male frog send advertisement pipe after refusal, or the refusal just sending when the male frog embraces the female frog pipes.
Step 2: piping of needs identification done to identical processing to obtain Spectral Characteristic value;
Step 3: calculate the pipe similarity degree of feature database the inside species of batrachia to be detected and batrachia by HMM, export the name title of the most similar a kind of batrachia, and export the most similar sex title and the status flag of seeking a spouse.
In the present embodiment, the same typing feature database of data that real example is correct.
Embodiment five
Batrachia is sought a spouse the phase, and the numerical value such as male tweeting sound cent shellfish, frequency, duration are all comparatively obvious, is therefore originally implemented on the basis of embodiment tetra-, only limits to male record in measured data and feature database to compare, and can further improve mensuration accuracy.
The present embodiment comprises:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
Repeating step A to step C, preserves various parameters, sets up the batrachia feature database that pipes; Each Spectral Characteristic value has characterized its corresponding object species characteristic, has also further characterized its corresponding sex character, with and Reproductive State feature.
Step 2: piping of needs identification done to identical processing to obtain Spectral Characteristic value;
Step 3: non-male record in shielding characteristic storehouse, by HMM, calculate the pipe similarity degree of feature database the inside species of batrachia to be detected and batrachia, export the name title of the most similar a kind of batrachia, and the most similar sex title and the status flag of seeking a spouse of output.
The male frog of the present embodiment utilization phase sounding characteristic of seeking a spouse, dwindles feature database scope, is conducive to improve the accuracy that kind is differentiated.By male data, researchist also can carry out reasoning to other situations of population simultaneously.
Step 4: the data typing feature database that real example is correct.
Embodiment six
On the basis of embodiment five, the present embodiment is further introduced the batrachia individual identification based on sound, thereby is conducive to the research to the frog of Different Individual under specific environment.
The present embodiment comprises:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
Repeating step A to step C, preserves various parameters, sets up the batrachia feature database that pipes; Each Spectral Characteristic value has characterized its corresponding object species characteristic, has also further characterized its corresponding sex character, with and Reproductive State feature; Each Spectral Characteristic value has characterized its corresponding individuality, and each individuality has respectively a unique numbering.
Step 2: when needs individual identification, individuality to be discriminated (this individuality gathered the feature that pipes before experiment) is carried out to sound collection, piping of needs identification done to identical processing to obtain Spectral Characteristic value.
Step 3: calculate the pipe similarity degree of feature database the inside species of batrachia to be detected and batrachia by HMM, differentiate its individuality, Output rusults is the numbering of this frog.
In the present embodiment, only by sound, just can well identify batrachia individuality, successfully overcome in batrachia research in the past, need to cut in toe or body and inject injury and the impact that the insertion type labeling methods such as label are brought the frog.
The present embodiment, by the continuous monitoring to a certain individuality, can be studied features such as individual mechanics, migration situation, life-spans.
Claims (8)
1. based on Spectral Characteristic, identify a method for batrachia individual information, comprise the following steps:
Step 1: set up feature database;
A, the signal characteristic that pipes of the pattern batrachia collecting is carried out to noise elimination;
B, the batrachia after noise reduction is piped, according to the time division that pipes of the frog, become a plurality of tablet sections that pipe;
C, the tablet section that pipes is carried out to MFCC processing, extract Spectral Characteristic value;
How many according to the classification of tested frog kind, repeating step A is to step C, sets up the batrachia feature database that pipes; In described feature database, each tested frog kind is to there being at least one record; In every record, comprise sample Spectral Characteristic value, each Spectral Characteristic value has characterized its corresponding object kind;
Step 2: process measured data;
By the piping through noise reduction, cut apart and be extracted as measured data of needs identification, obtain the Spectral Characteristic value of measured data;
Step 3: determination step;
By HMM, calculate the similarity degree of measured data and feature database the inside species, export name title and the Spectral Characteristic value thereof of the most similar a kind of batrachia.
2. the method for identifying batrachia individual information based on Spectral Characteristic as claimed in claim 1, is characterized in that: also comprise step 4: step 3 is augmented in feature database by the correct test data of real example.
3. the method for identifying batrachia individual information based on Spectral Characteristic as claimed in claim 1, it is characterized in that: in described step 3, all records in measured data and feature database are compared, the all records that are consistent with measured data in marker characteristic storehouse, in all records that are labeled, the highest attribute of the frequency of occurrences is judged as the attribute being consistent with measured data.
4. the method based on Spectral Characteristic evaluation batrachia individual information as described in claim 1,2 or 3, is characterized in that: in the record of described feature database, each Spectral Characteristic value has also characterized its corresponding Individual Size feature, individual health status flag; In described step 3, first judge frog kind, then judge measurand Individual Size feature, individual health status flag.
5. the method based on Spectral Characteristic evaluation batrachia individual information as described in claim 1,2 or 3, is characterized in that: in the record of described feature database, each Spectral Characteristic value has also characterized its corresponding sex information; In described step 3, also comprise the step of judging measurand sex character.
6. the method for identifying batrachia individual information based on Spectral Characteristic as claimed in claim 5, is characterized in that: in the record of described feature database, each Spectral Characteristic value has also characterized its corresponding Reproductive State information; In described step 3, also comprise the step of judging measurand Reproductive State.
7. the method for identifying batrachia individual information based on Spectral Characteristic as claimed in claim 6, is characterized in that: in described step 3, measured data only with feature database in sex be that male record is compared.
8. the method for identifying batrachia individual information based on Spectral Characteristic as claimed in claim 7, it is characterized in that: in the record of described feature database, the corresponding unique individuality numbering of each male, in described determination step, also comprises individual determination step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410238578.0A CN103985385A (en) | 2014-05-30 | 2014-05-30 | Method for identifying Batrachia individual information based on spectral features |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410238578.0A CN103985385A (en) | 2014-05-30 | 2014-05-30 | Method for identifying Batrachia individual information based on spectral features |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103985385A true CN103985385A (en) | 2014-08-13 |
Family
ID=51277328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410238578.0A Pending CN103985385A (en) | 2014-05-30 | 2014-05-30 | Method for identifying Batrachia individual information based on spectral features |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103985385A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392722A (en) * | 2014-11-28 | 2015-03-04 | 电子科技大学 | Sound based biotic population identification method and system |
| CN105139852A (en) * | 2015-07-30 | 2015-12-09 | 浙江图维电力科技有限公司 | Engineering machinery recognition method and recognition device based on improved MFCC (Mel Frequency Cepstrum Coefficient) sound features |
| CN107369451A (en) * | 2017-07-18 | 2017-11-21 | 北京市计算中心 | A kind of birds sound identification method of the phenology research of auxiliary avian reproduction phase |
| CN109599120A (en) * | 2018-12-25 | 2019-04-09 | 哈尔滨工程大学 | One kind being based on large-scale farming field factory mammal abnormal sound monitoring method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030125946A1 (en) * | 2002-01-03 | 2003-07-03 | Wen-Hao Hsu | Method and apparatus for recognizing animal species from an animal voice |
| CN101976564A (en) * | 2010-10-15 | 2011-02-16 | 中国林业科学研究院森林生态环境与保护研究所 | Method for identifying insect voice |
| CN103117061A (en) * | 2013-02-05 | 2013-05-22 | 广东欧珀移动通信有限公司 | Method and device for identifying animals based on voice |
| US20130282379A1 (en) * | 2012-04-24 | 2013-10-24 | Tom Stephenson | Method and apparatus for analyzing animal vocalizations, extracting identification characteristics, and using databases of these characteristics for identifying the species of vocalizing animals |
-
2014
- 2014-05-30 CN CN201410238578.0A patent/CN103985385A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030125946A1 (en) * | 2002-01-03 | 2003-07-03 | Wen-Hao Hsu | Method and apparatus for recognizing animal species from an animal voice |
| CN101976564A (en) * | 2010-10-15 | 2011-02-16 | 中国林业科学研究院森林生态环境与保护研究所 | Method for identifying insect voice |
| US20130282379A1 (en) * | 2012-04-24 | 2013-10-24 | Tom Stephenson | Method and apparatus for analyzing animal vocalizations, extracting identification characteristics, and using databases of these characteristics for identifying the species of vocalizing animals |
| CN103117061A (en) * | 2013-02-05 | 2013-05-22 | 广东欧珀移动通信有限公司 | Method and device for identifying animals based on voice |
Non-Patent Citations (10)
| Title |
|---|
| GERHARDT H C: ""Female mate choice in tree frogs: static and dynamic acoustic criteria"", 《ANIMAL BEHAVIOUR》 * |
| GUO MING ET AL: ""A study on the crying feature of crested ibis in different growing periods"", 《JOURNAL OF NORTHWEST UNIVERSITY》 * |
| MORRISON C ET AL: ""Mate selection in Litoris and Litoria xanthomera"", 《AUSTRAL ECOLOGY》 * |
| RAMER J, JENSSEN T, HURST C: ""Size-related variation in the advertisement call of Rana clamitans (Anura: Ranidae), and its effect on conspecific males"", 《COPERIA》 * |
| SULLIVAN B K, HINSHAW S H: ""Variation in the advertisement calls and male calling behavior in the spring peeper"", 《COPERIA》 * |
| 李佩珣等: ""繁殖期黄喉鹀的领域鸣唱及其种内个体识别"", 《动物学研究》 * |
| 王岩: ""基于动物叫声的物种识别技术的研究"", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 竺乐庆等: ""基于MFCC和GMM的昆虫声音自动识别"", 《昆虫学报》 * |
| 陈平: ""温度和体大小对棘胸蛙鸣声的影响及其地理变异"", 《中国优秀硕士学位论文全文数据库基础科学辑》 * |
| 黄英来: ""基于动物声音的个体辨认技术的研究"", 《中国优秀博硕士学位论文全文数据库（硕士）信息科技辑》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392722A (en) * | 2014-11-28 | 2015-03-04 | 电子科技大学 | Sound based biotic population identification method and system |
| CN105139852A (en) * | 2015-07-30 | 2015-12-09 | 浙江图维电力科技有限公司 | Engineering machinery recognition method and recognition device based on improved MFCC (Mel Frequency Cepstrum Coefficient) sound features |
| CN107369451A (en) * | 2017-07-18 | 2017-11-21 | 北京市计算中心 | A kind of birds sound identification method of the phenology research of auxiliary avian reproduction phase |
| CN109599120A (en) * | 2018-12-25 | 2019-04-09 | 哈尔滨工程大学 | One kind being based on large-scale farming field factory mammal abnormal sound monitoring method |
| CN109599120B (en) * | 2018-12-25 | 2021-12-07 | 哈尔滨工程大学 | Abnormal mammal sound monitoring method based on large-scale farm plant |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Knight et al. | Recommendations for acoustic recognizer performance assessment with application to five common automated signal recognition programs | |
| CN107577688B (en) | Original article influence analysis system based on media information acquisition | |
| US10403282B2 (en) | Method and apparatus for providing voice service | |
| CN102799605B (en) | A kind of advertisement detecting method and system | |
| Keen et al. | A comparison of similarity-based approaches in the classification of flight calls of four species of North American wood-warblers (Parulidae) | |
| CN108305615A (en) | A kind of object identifying method and its equipment, storage medium, terminal | |
| CN106782615A (en) | Speech data emotion detection method and apparatus and system | |
| WO2009090584A2 (en) | Method and system for activity recognition and its application in fall detection | |
| CN112712809B (en) | Voice detection method and device, electronic equipment and storage medium | |
| CN111859010B (en) | Semi-supervised audio event identification method based on depth mutual information maximization | |
| CN103985385A (en) | Method for identifying Batrachia individual information based on spectral features | |
| Adi et al. | Acoustic censusing using automatic vocalization classification and identity recognition | |
| Hagiwara et al. | BEANS: The benchmark of animal sounds | |
| CN115410711B (en) | White feather broiler health monitoring method based on sound signal characteristics and random forest | |
| Dong et al. | A novel representation of bioacoustic events for content-based search in field audio data | |
| CN115048984A (en) | Sow oestrus recognition method based on deep learning | |
| CN111462774B (en) | Music emotion credible classification method based on deep learning | |
| CN116318457B (en) | Radio signal monitoring method and system | |
| CN111583963A (en) | Method, device and equipment for detecting repeated audio and storage medium | |
| CN108520740B (en) | Audio content consistency analysis method and analysis system based on multiple characteristics | |
| CN114443930A (en) | News public opinion intelligent monitoring and analyzing method, system and computer storage medium | |
| KR102573186B1 (en) | Apparatus, method, and recording medium for providing animal sound analysis information | |
| CN114648688A (en) | Method, system and equipment for evaluating landscape level along high-speed rail and readable storage medium | |
| Kohlsdorf et al. | Feature Learning and Automatic Segmentation for Dolphin Communication Analysis. | |
| Wang et al. | A hierarchical birdsong feature extraction architecture combining static and dynamic modeling |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140813 |