CN101604330A - Method for automatically identifying circular microalgae based on shell face texture - Google Patents
Method for automatically identifying circular microalgae based on shell face texture Download PDFInfo
- Publication number
- CN101604330A CN101604330A CNA2009101122306A CN200910112230A CN101604330A CN 101604330 A CN101604330 A CN 101604330A CN A2009101122306 A CNA2009101122306 A CN A2009101122306A CN 200910112230 A CN200910112230 A CN 200910112230A CN 101604330 A CN101604330 A CN 101604330A
- Authority
- CN
- China
- Prior art keywords
- algae
- image
- circular
- microalgae
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
Based on the method for automatically identifying circular microalgae of shell face texture, relate to a kind of automatic distinguishing method for image.A kind of method for automatically identifying circular microalgae based on shell face texture with higher discrimination is provided.Set up the original little algae image data base and the computing machine of packing into microscope and image capture device; Adopt partition means, extract the frond region of target microalgae and generate the pure algae image A of 256 looks; A is carried out Fourier transform get spectral image B; With 0 the beginning, every 1/10 couple of B of algae image radius carry out ring-type sample radius spectrum signature sequence; Since 0 °, with every 10 ° to B radially sample up to 90 ° angle spectrum signature sequence, two sequence additions are got characteristic sequence to be identified; Record in characteristic sequence to be identified and the ALGAE CHARACTERISTICS sequence library is compared with definite wherein immediate record with editing distance by computing machine, and the optimal candidate kind of corresponding record as algae image to be identified.
Description
Technical field
The present invention relates to a kind of automatic distinguishing method for image, especially relate to and a kind ofly obtain spectral image by little algae micro-image is carried out Fourier transform operation, ring-type sampling and radially sampling by spectral image obtain the spectrum signature sequence; By to the spectrum signature sequence of algae to be detected and the comparison of spectrum signature sequence library, obtain the method for final classification results.
Background technology
Algae is the unicellular organism that a class is distributed widely in all kinds of habitats, and balance of ecosystem is played crucial effects.It is earth primary productivity of marine ecosystem chief component ([1] Mann DG.The species concept in diatoms.Phycologia, 1999,38:437-495).In addition, algae all is widely used in a lot of fields, as water quality detection ([2] Prygiel J, Coste M, Bukowska J.Review of the major diatom-based techniques for the qualityassessment of continental surface waters.In:Use of algae for monitoring rivers, Prygiel J, Coste M, Bukowska J. (eds), Agence de l ' Eau Artois-Picardie, Douai, France.1998:224-238; [3] Kelly MG.Water quality assessment by algal monitoring.In:The handbook of environmental monitoring, F.Burden, I.Mckelvie, A.Guenther and U.Fo:rstner (eds), McGraw-Hill, New York), culture, petroleum prospecting, nanometer technology ([4] Ryan WD, Richard G, Star Trek replicators and diatom nanotechnology[J] .Trends inBiotechnol, 2003,21:325-328) and red tide forecast etc., these use the evaluation work that all be unable to do without algae.
At present, evaluation of algae and quantitative test mostly are at microscopically, by what manually finish, need higher classification level professional technology, and be both consuming time, again effort.All seeking a kind of method for quickly identifying both at home and abroad, to satisfy the demand on algae is identified.
The profile that the impassable link that algae image is discerned automatically is a frond extracts.Yet traditional image processing method can't be done effective processing to algae image, and at first, the noise spot of micro-image is more, the template at traditional detection edge, and as Prewitt, Roberts, Sobel is affected by noise big; And gray level threshold segmentation is subjected to the very big influence of micro-image light source, and the directional problems of the lighting source that micro-image is common also is easy to generate the unbalanced effect of illumination, makes being divided into power and also can't meeting the demands of this method; In addition, the existence of large-scale impurity such as bubble makes that cutting apart subsequent processing steps can't realize intellectuality.Based on the automatic identification technology of algae image feature have conveniently, characteristics such as stable, directly perceived, become one both domestic and external gradually and study focus.
Because a large portion microalgae has different profile characteristics, at present, people mainly concentrate on according to algae appearance profile feature the automatic identification technology of algae and classify, in each category feature of algae image, contour feature be a kind of the most directly perceived, be easy to received algae recognition methods most.Yet, with the more closely-related type algae that swims of human being's production life in, appearance profile is very big for the proportion that circular algae accounts for, center guiding principle as diatom, its most of kind all has circular profile, for little algae of the type, it is obviously impossible only to classify by resemblance, can only be by the judgement of classifying of the texture feature of frond.
The inventor shows by a large amount of literature searches, though in some algae image automatic recognition systems, used textural characteristics to improve the accuracy of algae identification, yet, identification to the textural characteristics of algae mainly concentrates on (as the ADIAC engineering in Europe) on the non-circular diatom in the world, and circular algae be yet there are no report based on the automatic identifying method of texture.With regard to diatom, it adheres to different big classes separately its non-circular diatom (mainly being pennates) and circular diatom (mainly being the center guiding principle), and the compositing characteristic of its shell face texture also has very big-difference, and the texture recognition technology of non-circular algae is for circular algae and inapplicable.
Summary of the invention
The purpose of this invention is to provide a kind of method for automatically identifying circular microalgae based on shell face texture with higher discrimination.
The present invention includes following steps:
1) sets up the original little algae image data base of high definition and the computing machine of packing into microscope and image capture device;
2), extract the frond region of target microalgae and generate the pure algae image A of 256 looks by adopting partition means;
3) pure algae image A is carried out Fourier transform operation, get spectral image B;
4) with 0 beginning, carry out the ring-type sampling every 1/10 couple of spectral image B of algae image radius, obtain radius spectrum signature sequence; Since 0 °, spectral image B is radially sampled,, obtain angle spectrum signature sequence up to 90 ° every 10 °, with radius spectrum signature sequence and the addition of angle spectrum signature sequence, get characteristic sequence to be identified;
5) record in characteristic sequence to be identified and the ALGAE CHARACTERISTICS sequence library is compared with definite wherein immediate record with editing distance by computing machine, and the optimal candidate kind of corresponding record as algae image to be identified.
Described little algae preferably appearance profile is circular algae, and as the center guiding principle of diatom, the original little algae image of described high definition is that the resolution of the every secondary primitive algae image of requirement reaches 100 * 100 at least.
Described partition means can adopt computer picture dividing methods such as edge analysis or Threshold Segmentation.
The present invention based on principle be:
The texture feature of circular algae shows and mainly is radiativity on the direction, and the texture structure element mainly is the characteristics of hole line shape.Difference between the different circular algaes mainly is the difference of hole line size distribution and the difference of radiativity arrangement mode.Fourier spectrum is highly suitable for describing the directivity or the two-dimensional model of chain image, and it is very easy to find the characteristic of image in spatial domain, can detect the size and the spatial organization of texture primitive.For describing spectrum signature, by ring-type sampling and radially sampling, it can be reduced to the representation of one dimension, wherein the ring-type sampling has embodied the size distribution characteristic of algae texture structure element (as the Kong Wen of diatom), and radially sampling is corresponding with the periodicity of texture structure element.
Because the present invention mainly relies on the distinctive texture of circular algae to arrange feature, adopts a series of images treatment technology, obtains classification results, so the present invention not only has better stability and anti-interference, and has higher discrimination.
Description of drawings
Fig. 1 is the identification figure of the present invention to a width of cloth circular microalgae micro-image.Wherein (a) is the circular algae micro-single algae image of original image after gray processing is handled, (b) be cutting apart of figure a of back image, (c) spectral image that obtains through Fourier transform for figure b (d) is ring-type sampling characteristic sequence figure, (e) is ring-type sampling characteristic sequence figure.
Embodiment
Below in conjunction with embodiment technical scheme of the present invention is described in further detail.
1) take 10 kinds of micro-single algae images of circular algae under microscope and high precision digital camera, and it is carried out gray processing handle, Fig. 1 a provides the wherein circular algae micro-single algae image of 3 kinds of original images after gray processing is handled.
2) by adopting the rim detection cutting techniques, extract the frond region of target microalgae and generate the pure algae image of 256 looks, Fig. 1 b provides the result of image shown in Fig. 1 a after cutting apart.
3) obtain spectral image by Fourier transform operation, Fig. 1 c provides the spectral image that Fig. 1 b obtains through Fourier transform.
4) with 0 beginning, carry out ring-type sampling every 1/10 couple of spectral image B of length algae image radius, obtain length and be 10 radius spectrum signature sequence, ring-type sampling characteristic sequence figure is referring to Fig. 1 d.Since 0 °,,, obtain angle and be 10 ° angle spectrum signature sequence up to 90 ° it is radially sampled every 10 °.Two sequence additions obtain characteristic sequence, and ring-type sampling characteristic sequence figure is referring to Fig. 1 e.
5) set up the characteristic sequence database.
6) take the micro-digital micrograph image of algae to be identified, obtain the characteristic sequence of algae to be identified according to the method from the step 1) to the step 4).
7) immediate record in calculating and the characteristic sequence database is its optimal candidate identity as algae to be identified.
Claims (5)
1. based on the method for automatically identifying circular microalgae of shell face texture, it is characterized in that may further comprise the steps:
1) sets up the original little algae image data base of high definition and the computing machine of packing into microscope and image capture device;
2), extract the frond region of target microalgae and generate the pure algae image A of 256 looks by adopting partition means;
3) pure algae image A is carried out Fourier transform operation, get spectral image B;
4) with 0 beginning, carry out the ring-type sampling every 1/10 couple of spectral image B of algae image radius, obtain radius spectrum signature sequence; Since 0 °, spectral image B is radially sampled,, obtain angle spectrum signature sequence up to 90 ° every 10 °, with radius spectrum signature sequence and the addition of angle spectrum signature sequence, get characteristic sequence to be identified;
5) record in characteristic sequence to be identified and the ALGAE CHARACTERISTICS sequence library is compared with definite wherein immediate record with editing distance by computing machine, and the optimal candidate kind of corresponding record as algae image to be identified.
2. the method for automatically identifying circular microalgae based on shell face texture as claimed in claim 1 is characterized in that described little algae is that appearance profile is circular algae.
3. the method for automatically identifying circular microalgae based on shell face texture as claimed in claim 2 is characterized in that described appearance profile is the center guiding principle of diatom for circular algae.
4. the method for automatically identifying circular microalgae based on shell face texture as claimed in claim 1 is characterized in that the original little algae image of described high definition, is that the resolution of the every secondary primitive algae image of requirement reaches 100 * 100 at least.
5. the method for automatically identifying circular microalgae based on shell face texture as claimed in claim 1 is characterized in that described partition means adopts the computer picture dividing method of edge analysis or Threshold Segmentation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910112230A CN101604330B (en) | 2009-07-17 | 2009-07-17 | Shell surface texture method for automatically identifying circular microalgae |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910112230A CN101604330B (en) | 2009-07-17 | 2009-07-17 | Shell surface texture method for automatically identifying circular microalgae |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101604330A true CN101604330A (en) | 2009-12-16 |
| CN101604330B CN101604330B (en) | 2012-09-05 |
Family
ID=41470062
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910112230A Expired - Fee Related CN101604330B (en) | 2009-07-17 | 2009-07-17 | Shell surface texture method for automatically identifying circular microalgae |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101604330B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103530647A (en) * | 2013-10-10 | 2014-01-22 | 哈尔滨工程大学 | Texture classification method on basis of fractional Fourier transform (FrFT) |
| CN107679509A (en) * | 2017-10-19 | 2018-02-09 | 广东工业大学 | A kind of small ring algae recognition methods and device |
| CN110543580A (en) * | 2019-08-27 | 2019-12-06 | 安徽生物工程学校 | Picture comparison and identification method for common algae |
-
2009
- 2009-07-17 CN CN200910112230A patent/CN101604330B/en not_active Expired - Fee Related
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103530647A (en) * | 2013-10-10 | 2014-01-22 | 哈尔滨工程大学 | Texture classification method on basis of fractional Fourier transform (FrFT) |
| CN103530647B (en) * | 2013-10-10 | 2017-02-08 | 哈尔滨工程大学 | Texture classification method on basis of fractional Fourier transform (FrFT) |
| CN107679509A (en) * | 2017-10-19 | 2018-02-09 | 广东工业大学 | A kind of small ring algae recognition methods and device |
| CN107679509B (en) * | 2017-10-19 | 2021-04-02 | 广东工业大学 | Cyclotella tenera identification method and device |
| CN110543580A (en) * | 2019-08-27 | 2019-12-06 | 安徽生物工程学校 | Picture comparison and identification method for common algae |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101604330B (en) | 2012-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Improved kiwifruit detection using pre-trained VGG16 with RGB and NIR information fusion | |
| CN104112370B (en) | Parking lot based on monitoring image intelligent car position recognition methods and system | |
| CN106557764B (en) | A kind of water level recognition methods based on binary-coded character water gauge and image procossing | |
| CN108764154B (en) | Water surface garbage identification method based on multi-feature machine learning | |
| CN106446894B (en) | A method of based on outline identification ball-type target object location | |
| CN106934386B (en) | A kind of natural scene character detecting method and system based on from heuristic strategies | |
| CN108765402B (en) | Non-woven fabric defect detection and classification method | |
| CN108596038B (en) | Method for identifying red blood cells in excrement by combining morphological segmentation and neural network | |
| CN111598856B (en) | Chip surface defect automatic detection method and system based on defect-oriented multipoint positioning neural network | |
| CN101877074A (en) | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics | |
| CN103514448A (en) | Method and system for navicular identification | |
| CN102654464A (en) | Copper strip surface defect detection system based on multi-feature fuzzy recognition | |
| CN110108712A (en) | Multifunctional visual sense defect detecting system | |
| CN105678735A (en) | Target salience detection method for fog images | |
| CN103177266A (en) | Intelligent stock pest identification system | |
| WO2021109011A1 (en) | Intelligent capacitor internal defect detection method based on ultrasound image | |
| CN110443278A (en) | A kind of detection method, device and the equipment of solar battery sheet grid line thickness exception | |
| CN101604330B (en) | Shell surface texture method for automatically identifying circular microalgae | |
| CN109858386A (en) | A kind of microalgae cell recognition methods based on fluorescence microscope images | |
| CN102073872A (en) | Image-based method for identifying shape of parasite egg | |
| CN104573662B (en) | A kind of cloud sentences method and system | |
| CN110211128B (en) | Loess plateau terrace extraction method based on remote sensing image and DEM | |
| CN114972204A (en) | Steel product surface crack detection method and equipment | |
| CN101694721B (en) | Method for locating circular algae in microimage | |
| Alekhin et al. | Optical-electronic system for express analysis of mineral raw materials dressability by color sorting 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 | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120905 Termination date: 20200717 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |