CN102156710A - Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method - Google Patents

Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method Download PDF

Info

Publication number
CN102156710A
CN102156710A CN 201110048981 CN201110048981A CN102156710A CN 102156710 A CN102156710 A CN 102156710A CN 201110048981 CN201110048981 CN 201110048981 CN 201110048981 A CN201110048981 A CN 201110048981A CN 102156710 A CN102156710 A CN 102156710A
Authority
CN
China
Prior art keywords
cloud model
membership
evaluation
plant
degree
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
CN 201110048981
Other languages
Chinese (zh)
Inventor
彭琳
刘宗田
杨林楠
钟飞
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.)
University of Shanghai for Science and Technology
Original Assignee
University of Shanghai for Science and Technology
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 University of Shanghai for Science and Technology filed Critical University of Shanghai for Science and Technology
Priority to CN 201110048981 priority Critical patent/CN102156710A/en
Publication of CN102156710A publication Critical patent/CN102156710A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a plant identification method based on a cloud model and a TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method. The method comprises the following steps: constructing a plant shape feature specimen database; utilizing a trapezium-cloud model to compare the shape features of a plant to be identified with the plant shape feature specimen database to acquire a comparative membership between the plant to be identified and the shape feature specimen database, thus completing the primary identification of the plant to be identified; when a plurality of identification results exist, utilizing a normal cloud model to carry out accurate matching calculation on the retrieval results so as to acquire a comparative accurate membership between the plant to be identified and the shape feature specimen database; and comprehensively evaluating the membership by utilizing the TOPSIS method to identify the plant. The method can comprehensively evaluate the final identification result by adopting the TOPSIS method, and can completely, reasonably and accurately carry out advantage and disadvantage sequencing according to certain evaluation indexes, so that the evaluation process is clear, and the evaluation result is objective.

Description

A kind of plant discrimination method based on cloud model and TOPSIS method
Technical field
The invention belongs to plant authentication technique field, specifically relate to a kind of plant discrimination method based on cloud model and TOPSIS method.
Background technology
Important effect is being brought into play in the agricultural data storehouse in the development and use of agricultural information resources, agricultural organization has built large quantities of agricultural datas storehouse both at home and abroad at present, wherein, just there are AGRIS, IFIS, AGRI-COLA, CABI four multi-form agriculture databases in foremost in the world agricultural data storehouse system, it is the agricultural system database (AGRIS) of FAO (Food and Agriculture Organization of the United Nation), international food information database (IFIS), United States Department of Agriculture's agricultural on-line access database (AGRI-COLA) and international agro-ecology central database (CAB1).The representational plant database of China just has: China national Agricultural Science Data Sharing center (www.agridata.cn), Chinese digital plant specimen shop (http://www.cvh.org.cn), Chinese Plants database (www.plant.csdb.cn), Chinese Plants science net (www.chinaplant.org), Chinese biological diversity information system (bd.brim.ac.cn), Chinese Plants image library (www.plantphoto.cn) etc.
These databases have comprised abundant professional knowledge, stored the relevant scientific and technical information of a large amount of agriculturals, because these databases are strongly professional, and can only adopt the keyword retrieval mode to inquire about, increased the operation easier of operator to database, reduce the service efficiency of database, caused the wasting of resources.It mainly shows as: (1) requires height to the operator.Use the operator of these databases must expertly use computing machine, understand the database retrieval interface, grasp search strategy.Simultaneously, must be to theme, keyword, mechanism, in full, autograph etc.-as retrieval concept and search channel to have gained some understanding and grasp.But the user of plant database not only comprises domain expert and the agriculture technical staff that strong professional knowledge is arranged, and also comprises peasant and non-agricultural scientific worker, and these most of personnel are difficult to import accurately as requested term; (2) keyword retrieval mode requires to retrieve enquirement must be in strict accordance with the form input of regulation, have only when assembly is mated fully and just can obtain Search Results, this retrieval mode of puing question to sign to be consistent with retrieval on literal, very difficult realization in terms of content with the conceptive result for retrieval of meeting consumers' demand that retrieves, will cause the recall ratio of result for retrieval and precision ratio lower.
At top problem, some researchers utilize digital image processing techniques to realize the plant discriminating, according to identification result database are inquired about, thereby are obtained relevant professional knowledge.
But also there is following problem in this technology at present: (1) present image acquisition is passed through scanner substantially, and background is simple, and image segmentation is relative simple with description, is difficult to handle complex background; (2) image data base of different researchers' uses has nothing in common with each other, and is difficult to the relatively performance quality of recognition effect, and is difficult to use in practice; (3) current existing image classification system can only handle usually that several scope is less to hundred kind of plant, is difficult to realize that extensive, multi-class plant differentiates.Yet more existing researchers propose:
Bibliographical information is arranged, and its exercise question is: " uncertainty in the representation of knowledge " (this article author is Li Deyi, and " the Chinese engineering science " the 2nd that is published in publication in 2000 rolled up the 73rd～79 page of the 10th phase).It is that it comprises with certain qualitativing concept of natural language value representation and the uncertain transformation model between its quantificational expression that this article discloses cloud model:
(1) definition of cloud
If U is a quantitative domain with the perfect number value representation, C is the qualitativing concept on the U, if quantitative values
Figure 2011100489813100002DEST_PATH_IMAGE001
∈ U, and
Figure 564413DEST_PATH_IMAGE001
Be once realizing at random of qualitativing concept C, Degree of certainty to C
Figure 538896DEST_PATH_IMAGE002
∈ [O, 1] is the random number that steady tendency is arranged: : U → [O, 1],
Figure 975563DEST_PATH_IMAGE004
,
Figure 2011100489813100002DEST_PATH_IMAGE005
Then
Figure 210104DEST_PATH_IMAGE001
Distribution on domain is called cloud model, abbreviates cloud as, each Be called a water dust.
(2) numerical characteristic of cloud
The numerical characteristic of cloud model is with expecting
Figure 842784DEST_PATH_IMAGE006
, entropy
Figure 2011100489813100002DEST_PATH_IMAGE007
With super entropy
Figure 198986DEST_PATH_IMAGE008
Characterize, reflected the overall permanence of qualitativing concept C.
Being the expectation of water dust in the domain space distribution, is exactly the point that can represent qualitativing concept, has reflected the water dust group's of this notion cloud center of gravity.
Figure 768693DEST_PATH_IMAGE007
Be the uncertainty measure of qualitativing concept, determine jointly, disclosed the relevance of ambiguity and randomness by the randomness and the ambiguity of notion.Entropy
Figure 185768DEST_PATH_IMAGE007
Be the tolerance of qualitativing concept randomness on the one hand, reflected the dispersion degree that to represent the water dust of this qualitativing concept; Be again the tolerance of qualitativing concept ambiguity on the other hand, reflected in the domain space span of the water dust that can be accepted by notion.With
Figure 170429DEST_PATH_IMAGE007
This numerical characteristic reflects ambiguity and randomness simultaneously, has also embodied the relevance between the two.
Figure 989350DEST_PATH_IMAGE008
Be probabilistic tolerance of entropy, reflected the coherency of representing the uncertainty of being had a few of this notion in the number field space, be i.e. the condensation degree of water dust.Its size has been represented the dispersion degree and the thickness of cloud indirectly, is determined jointly by the randomness and the ambiguity of entropy.
The numerical characteristic of cloud is just to delineate the whole cloud that is made of thousands of water dust with three numerical value, and ambiguity and randomness in the language value of qualitative representation are fully integratible into together.Because specific implementation method difference constitutes dissimilar clouds, as normal cloud model, trapezoidal cloud model, half cloud model etc.Wherein, normal cloud model and trapezoidal cloud model.
Also have the books report, its title is " Multiple Attribute Decision Making:Methods and Application " (this book author is: C.L.Hwang and Yoon K., the publication in 1981 years of Berlin Springe publishing house).A kind of method that sorts according to the degree of closeness of limited evaluation object and idealized target is disclosed in this book---discrete type approaches ideal solution ranking method (Technique for Order Preference by Similarity to Ideal Solution, abbreviation TOPSIS).The raw data matrix of its principle after based on normalization, find out optimal case and space of the most bad forecast scheme configuration in the limited scheme, certain object to be evaluated can be considered a point on this space, can obtain distance (the Euclidean distance commonly used between this point and optimal case and the most bad scheme in view of the above, claim Euclidean distance again), thereby draw the relative degree of closeness of this object and optimal case, take this to carry out the evaluation of scheme quality.
Summary of the invention
The problem and shortage that exists of prior art in view of the above, the technical problem to be solved in the present invention provides a kind of plant discrimination method based on cloud model and TOPSIS method, this method can retrieve by measuring plants from the deme various plants or from the database of preserving a large amount of plant specimens quickly and easily, realizes the discriminating to plant.
In order to address the above problem, the present invention adopts following technical proposals:
A kind of plant discrimination method based on cloud model and TOPSIS method is characterized in that this method has at first made up the resemblance sample database of plant; Utilize trapezoidal cloud model to be compared then, obtain the degree of membership compared with resemblance sample database by measuring plants, realized by the preliminary discriminating of measuring plants by the external appearance characteristic sample database of the external appearance characteristic of measuring plants and plant; When identification result when being a plurality of, utilize normal cloud model that result for retrieval is accurately mated calculating again, obtain the accurate degree of membership of being compared with resemblance sample database by measuring plants; Utilize the TOPSIS method that degree of membership is carried out comprehensive evaluation at last, identify plant, concrete steps are as follows:
(1), makes up the resemblance sample database of plant;
(2), utilize trapezoidal cloud model, calculate by the degree of membership of measuring plants, tentatively differentiate by measuring plants;
(3), judge whether to be less than or equal to 1 by the degree of membership of measuring plants, if be to be less than or equal to 1, then change step (5) by the degree of membership of measuring plants, if be to be less than or equal to 1 to equal 1, then change step (4) by the degree of membership of measuring plants;
(4), utilize normal cloud model, be that 1 sample calculates with carried out normal cloud model by measuring plants to degree of membership, obtain by the accurate degree of membership of measuring plants;
(5), utilize the TOPSIS method, degree of membership is carried out comprehensive evaluation identifies by measuring plants.
The resemblance sample database of structure plant described in the above-mentioned steps (1), these sample data stock is put to can be in order to distinguish the resemblance of floristic plant, and its resemblance value is the numeric type data.
Utilize trapezoidal cloud model described in the above-mentioned steps (2), calculate, differentiate that tentatively its operation steps is as follows by measuring plants by the degree of membership of measuring plants:
(21), determine the expectation interval of trapezoidal cloud model
Figure 2011100489813100002DEST_PATH_IMAGE009
: according to the resemblance span in the resemblance sample storehouse of plant, determine the expectation interval of trapezoidal cloud model
Figure 890178DEST_PATH_IMAGE009
, wherein
Figure 63058DEST_PATH_IMAGE010
Be the lower limit of this resemblance,
Figure 2011100489813100002DEST_PATH_IMAGE011
Higher limit for this resemblance value;
(22), calculate trapezoidal cloud model entropy : utilize the rising cloud and the fall cloud of trapezoidal cloud model, the span of the resemblance of plant is expanded, be set at [3 between the expansion area
Figure 380962DEST_PATH_IMAGE012
,+3 ];
(23), the numerical characteristic that utilizes trapezoidal cloud model is to being described by the resemblance of measuring plants: determine trapezoidal cloud expectation curve equation by trapezoidal cloud expectation and entropy:
Figure 2011100489813100002DEST_PATH_IMAGE013
Figure 112868DEST_PATH_IMAGE014
Wherein,
Figure 2011100489813100002DEST_PATH_IMAGE015
The degree of membership of ordering for trapezoidal cloud x,
Figure 575684DEST_PATH_IMAGE016
For expecting interval lower limit,
Figure 41300DEST_PATH_IMAGE017
For expecting the interval upper limit,
Figure 424877DEST_PATH_IMAGE012
Entropy for trapezoidal cloud;
(24), utilize trapezoidal cloud model, calculate by the degree of membership of measuring plants: will be utilized trapezoidal cloud model to be analyzed respectively by the resemblance value in all resemblances of measuring plants and the sample storehouse, and obtain by each resemblance of measuring plants
Figure 366157DEST_PATH_IMAGE015
Utilizing normal cloud model described in the above-mentioned steps (4), is that 1 sample calculates with carried out normal cloud model by measuring plants to degree of membership, obtains by the accurate degree of membership of measuring plants, and its operation steps is as follows:
(41), determine the expectation value of normal cloud model:, determine the expectation value of normal cloud model according to profile eigenwert in the resemblance sample storehouse of plant
Figure 732198DEST_PATH_IMAGE018
,
Figure 943998DEST_PATH_IMAGE018
Intermediate value for the resemblance interval value;
(42), calculate the entropy of normal cloud model
Figure 50539DEST_PATH_IMAGE012
, entropy
Figure 717931DEST_PATH_IMAGE012
Calculating formula be:
(43), can determine normal state cloud expectation curve equation, this curvilinear equation is by normal state cloud expectation and entropy:
Figure 979070DEST_PATH_IMAGE020
Wherein,
Figure 438870DEST_PATH_IMAGE015
The degree of membership of ordering for trapezoidal cloud x,
Figure 292425DEST_PATH_IMAGE021
Be expectation value,
Figure 498803DEST_PATH_IMAGE012
Entropy for trapezoidal cloud;
(44), will to be calculated by the degree of membership of measuring plants by all resemblances of measuring plants and step (2) be that 1 sample is analyzed, and obtains by each resemblance of measuring plants
Figure 121415DEST_PATH_IMAGE015
Utilize the TOPSIS method described in the above-mentioned steps (5), degree of membership is carried out comprehensive evaluation identify by measuring plants, its operation steps is as follows:
(51), establish the evaluation matrix F of degree of membership comprehensive evaluation: if the degree of membership value that step (2) is calculated has only one or do not have one to equal at 1 o'clock, the evaluation matrix F is the degree of membership matrix of step (2) gained; Otherwise the evaluation matrix F is that step (3) is calculated the gained result;
(52), determine to estimate the ideal point of matrix F, the calculating formula of its ideal point is:
Figure 752116DEST_PATH_IMAGE022
Wherein,
Figure 30651DEST_PATH_IMAGE023
Be desirable point set,
Figure 37790DEST_PATH_IMAGE024
Be that i is individual by j value of evaluation of programme,
Figure 517837DEST_PATH_IMAGE025
Be the objective function numbering collection of asking maximum,
Figure 585019DEST_PATH_IMAGE026
It is the collection of functions of asking minimum target;
(53), determine to estimate the most not good enough of matrix F, its most not good enough calculating formula is:
Figure 147588DEST_PATH_IMAGE027
Wherein, Be the most not good enough collection,
Figure 292972DEST_PATH_IMAGE029
Be that i is individual by j value of evaluation of programme,
Figure 265476DEST_PATH_IMAGE025
Be the objective function numbering collection of asking maximum,
Figure 315340DEST_PATH_IMAGE026
It is the collection of functions of asking minimum target;
(54), calculate to estimate that each is by the distance of evaluation of programme to ideal point in the matrix F, its calculating formula is:
Figure 601965DEST_PATH_IMAGE030
，
Figure 44886DEST_PATH_IMAGE031
Wherein,
Figure 453870DEST_PATH_IMAGE032
Be that i is individual by the distance of evaluation of programme to ideal point, Be that i is individual by j value of evaluation of programme,
Figure 878084DEST_PATH_IMAGE033
Be the value of the j item of ideal point, n is by the number of evaluation of programme;
(55), calculate to estimate in the matrix F each by evaluation of programme to the most not good enough distance, its calculating formula is:
Figure 187231DEST_PATH_IMAGE034
，
Wherein,
Figure 791573DEST_PATH_IMAGE035
Being that i is individual is arrived the most not good enough distance by evaluation of programme,
Figure 482317DEST_PATH_IMAGE029
Be that i is individual by j value of evaluation of programme,
Figure 645970DEST_PATH_IMAGE036
Be the value of the most not good enough j item, n is by the number of evaluation of programme;
(56), calculate to estimate that each is by the connect recency of evaluation of programme to ideal point in the matrix F, its calculating formula is:
Figure 396757DEST_PATH_IMAGE037
，
Figure 908510DEST_PATH_IMAGE031
Wherein,
Figure 75049DEST_PATH_IMAGE038
Be that i is individual by the connect recency of evaluation of programme to ideal point,
Figure 355858DEST_PATH_IMAGE039
Being that i is individual is arrived the most not good enough distance by evaluation of programme, Be i by the distance of evaluation of programme to ideal point, n is by the number of evaluation of programme;
(57), according to estimating in the matrix F respectively by the recency that connects of evaluation of programme to ideal point
Figure 13946DEST_PATH_IMAGE038
Each case is made good and bad ordering.
The trapezoidal cloud model entropy of the described calculating of above-mentioned steps (22)
Figure 312072DEST_PATH_IMAGE012
Its spreading range is 20% of a resemblance lower limit, then trapezoidal cloud model entropy
Figure 447387DEST_PATH_IMAGE012
Calculating formula is: entropy
Figure 554625DEST_PATH_IMAGE012
=( * 0.2)/3.
A kind of plant discrimination method based on cloud model and TOPSIS method of the present invention has following effect compared with prior art:
(1), this method utilizes the outer property eigenwert of plant that unknown plant is differentiated, when the user has avoided the keyword retrieval plant database effectively, operator's professional knowledge is required high and result for retrieval and the unmatched problem of demand, enlarge range of search, improved retrieval precision;
(2), this method utilizes cloud model that the resemblance information of plant is carried out the digitizing description, realized the uncertain conversion between the qualitative and quantitative of resemblance information of plant, simultaneously, utilize the different digital characteristic of trapezoidal cloud model and normal cloud model that the plant resemblance is described respectively, guaranteeing to differentiate under the condition of precision, enlarge range of search, improved identification result;
(3), this method adopts TOPSIS that last identification result is carried out comprehensive evaluation, can be comprehensively, rationally, exactly certain several evaluation index is carried out quality and sorts, evaluation procedure is clear, evaluation result is objective.
Description of drawings
Fig. 1 is the process flow diagram of a kind of plant discrimination method based on cloud model and TOPSIS method of the present invention;
Fig. 2 is the synoptic diagram of trapezoidal cloud model;
Fig. 3 is the synoptic diagram of normal cloud model;
Fig. 4 is the resemblance sample database of bamboo;
Fig. 5 is a trapezoidal cloud model expectation value interval table (having saved bamboo kind title);
Fig. 6 is a resemblance entropy table;
Fig. 7 is the tested bamboo and the corresponding degree of membership of sample database resemblance based on ladder table cloud model;
Fig. 8 is the normal cloud model expectation value
Figure 346050DEST_PATH_IMAGE021
Table;
Fig. 9 is the normal cloud model entropy
Figure 601450DEST_PATH_IMAGE012
The value table;
Figure 10 is the tested bamboo and the corresponding degree of membership of sample database resemblance based on normal cloud model.
Embodiment
The present invention is described in further detail below in conjunction with the drawings and specific embodiments.
Present embodiment is with unknown plant---and bamboo is as differentiating that thing is embodiment.
With reference to Fig. 1, the plant discrimination method based on cloud model and TOPSIS method of the present invention, its step is as follows:
(1), makes up the resemblance sample database of plant
These sample data stock is put to can be in order to distinguish floristic plant resemblance, and its resemblance value is the numeric type data, and choosing 6 numerical value shape parameters such as the long and Ye Kuan of stalk height, diameter, internode length, stalk wall thickness, the leaf of bamboo is the resemblance parameter; From " world bamboo rattan ", select 15 kinds of common bamboos then, set up the resemblance sample database of bamboo., referring to Fig. 4.
(2), utilize trapezoidal cloud model, calculate by the degree of membership of measuring plants, realize that its operation steps is as follows by the preliminary discriminating of measuring plants:
(21), determine the expectation interval of trapezoidal cloud model
Figure 867871DEST_PATH_IMAGE009
: according to the data of the resemblance sample database table 1 of bamboo, determine the expectation value interval , for example, expect that interval value for interval [1～5], promptly is expressed as, =1, =5, transformation result as shown in Figure 5.
(22) calculate trapezoidal cloud model entropy
Figure 435195DEST_PATH_IMAGE012
Utilize the rising cloud and the fall cloud of trapezoidal cloud model, the span of the resemblance of plant is expanded, be set at [3 between the expansion area
Figure 630553DEST_PATH_IMAGE012
,+3
Figure 870910DEST_PATH_IMAGE012
], spreading range is 20% of a resemblance lower limit, so trapezoidal cloud model entropy
Figure 835324DEST_PATH_IMAGE012
Calculating formula: entropy
Figure 443548DEST_PATH_IMAGE012
=(
Figure 798306DEST_PATH_IMAGE041
* 0.2)/3.Calculate each respectively by the entropy of the resemblance of measuring plants
Figure 780037DEST_PATH_IMAGE012
, its result as shown in Figure 6.
(23), the numerical characteristic that utilizes trapezoidal cloud model is to being described by the resemblance of measuring plants: determine trapezoidal cloud expectation curve equation by trapezoidal cloud expectation and entropy:
Figure 598957DEST_PATH_IMAGE013
Figure 366363DEST_PATH_IMAGE014
Wherein, The degree of membership of ordering for trapezoidal cloud x;
Figure 56156DEST_PATH_IMAGE045
For expecting interval lower limit;
Figure 729583DEST_PATH_IMAGE047
For expecting the interval upper limit;
Figure 945188DEST_PATH_IMAGE012
Entropy for trapezoidal cloud.
For example, carry out trapezoidal cloud model to first in the resemblance sample database of bamboo and describe, the stalk height of bamboo is 1～5m, the resemblance of this bamboo, and the numerical characteristic of describing with trapezoidal cloud model is:
Figure 336855DEST_PATH_IMAGE048
Figure 394810DEST_PATH_IMAGE049
The synoptic diagram of this trapezoidal cloud model can be with reference to Fig. 2, and X-axis X is expressed as the stalk height of bamboo among Fig. 2, and unit is a rice; Y-axis
Figure 188323DEST_PATH_IMAGE003
Be expressed as degree of membership
Figure 574829DEST_PATH_IMAGE043
Wherein,
Figure 125896DEST_PATH_IMAGE045
For expecting interval lower limit,
Figure 253121DEST_PATH_IMAGE050
For expecting the interval upper limit,
Figure 901140DEST_PATH_IMAGE012
Entropy for trapezoidal cloud.
(24) utilize trapezoidal cloud model, calculate by the degree of membership of measuring plants: according to the expectation value interval of being calculated
Figure 190039DEST_PATH_IMAGE009
And entropy
Figure 559228DEST_PATH_IMAGE012
, utilize that trapezoidal cloud describes by the numerical characteristic of measuring plants, calculate the degree of membership of each resemblance of tested bamboo respectively, as shown in Figure 7.
(3), judge whether to be less than or equal to 1 by the degree of membership of measuring plants, if be to be less than or equal to 1, then change step (5) by the degree of membership of measuring plants, if be to be less than or equal to 1 to equal 1, then change step (4) by the degree of membership of measuring plants;
(4), utilize normal cloud model, degree of membership is 1 sample and is carried out normal cloud model by measuring plants and calculate, through step (2) obtain tested bamboo and 8,11 and the degree of membership that contrasts of 14 3 kind of bamboo be 1, utilize normal cloud model that tested bamboo and 8,11 and 14 is differentiated that its concrete steps are as follows:
(41) determine the expectation value of normal cloud model
According to profile eigenwert in the resemblance sample storehouse of plant, determine the expectation value of normal cloud model
Figure 224565DEST_PATH_IMAGE052
,
Figure 727090DEST_PATH_IMAGE052
Be the intermediate value of resemblance interval value, obtain the expectation value of normal cloud model, as shown in Figure 8.
(42) entropy of calculating normal cloud model
Figure 452469DEST_PATH_IMAGE012
, its calculating formula is: , according to [3
Figure 524122DEST_PATH_IMAGE012
,+3
Figure 146733DEST_PATH_IMAGE012
] between element surpassed 99.73%, its entropy
Figure 91948DEST_PATH_IMAGE012
Result of calculation as shown in Figure 9.
(43), can determine normal state cloud expectation curve equation, this curvilinear equation is by normal state cloud expectation and entropy:
Figure 996582DEST_PATH_IMAGE053
Wherein,
Figure 496397DEST_PATH_IMAGE043
The degree of membership of ordering for trapezoidal cloud x;
Figure 350345DEST_PATH_IMAGE052
Be expectation value;
Figure 653413DEST_PATH_IMAGE012
Entropy for trapezoidal cloud;
For example the stalk height of bamboo is 1～5m, the resemblance of this bamboo, and the numerical characteristic of describing with trapezoidal cloud is:
Figure 655129DEST_PATH_IMAGE054
The synoptic diagram of this normal cloud model can be with reference to Fig. 3, and X-axis X is expressed as the stalk height of bamboo among Fig. 3, and unit is a rice; Y-axis Be expressed as degree of membership
Figure 877479DEST_PATH_IMAGE043
Wherein,
Figure 617027DEST_PATH_IMAGE021
Be expectation value,
Figure 218954DEST_PATH_IMAGE012
Entropy for trapezoidal cloud.
(44), will to be calculated by the degree of membership of measuring plants by all resemblances of measuring plants and step (1) be that 1 sample is analyzed, and obtains by each resemblance of measuring plants
Figure 882410DEST_PATH_IMAGE043
According to the expectation value of being calculated
Figure 773268DEST_PATH_IMAGE021
And entropy , the numerical characteristic that utilizes trapezoidal cloud to describe calculates the degree of membership of each resemblance of tested bamboo respectively, and the result is as shown in figure 10.
(5), utilize the TOPSIS method, degree of membership is carried out comprehensive evaluation identifies by measuring plants, its concrete steps are as follows:
(51), establish the evaluation matrix F of degree of membership comprehensive evaluation
If the degree of membership that step (2) is calculated has 8,11,14 3 values to be 1, so the evaluation matrix F is the result of calculation of step (3), promptly
Figure 346778DEST_PATH_IMAGE056
(52), determine to estimate the ideal point of matrix F, the calculating formula of its ideal point is:
Figure 63193DEST_PATH_IMAGE022
Wherein, be desirable point set,
Figure 870874DEST_PATH_IMAGE058
J is an objective function numbering collection of asking maximum by j value of evaluation of programme to be i,
Figure 815872DEST_PATH_IMAGE026
Be the collection of functions of asking minimum target, its result of calculation is:
Figure 780942DEST_PATH_IMAGE059
(53), determine to estimate the most not good enough of matrix F, its most not good enough calculating formula is:
Wherein, Be the most not good enough collection,
Figure 839922DEST_PATH_IMAGE060
J is an objective function numbering collection of asking maximum by j value of evaluation of programme to be i,
Figure 525244DEST_PATH_IMAGE026
Be the collection of functions of asking minimum target, if its calculating is:
Figure 699305DEST_PATH_IMAGE061
(54), calculate to estimate that each is by the distance of evaluation of programme to ideal point in the matrix F, its calculating formula is:
Figure 543896DEST_PATH_IMAGE030
，
Wherein,
Figure 966815DEST_PATH_IMAGE063
Be that i is individual by the distance of evaluation of programme to ideal point,
Figure 931186DEST_PATH_IMAGE064
Be j value of i scheme,
Figure 817233DEST_PATH_IMAGE066
Be the value of the j item of ideal point, n is by the evaluation of programme number, and its result of calculation is:
Figure 425936DEST_PATH_IMAGE067
(55), calculate to estimate in the matrix F each by evaluation of programme to the most not good enough distance, its calculating formula is:
Figure 833653DEST_PATH_IMAGE034
，
Wherein,
Figure 803587DEST_PATH_IMAGE069
Being that i is individual is arrived the most not good enough distance by evaluation of programme,
Figure DEST_PATH_IMAGE070
Be j value of i scheme,
Figure 933656DEST_PATH_IMAGE072
Be the value of the most not good enough j item, n is by the evaluation of programme number, and its result of calculation is:
(56), calculate to estimate that each is by the connect recency of evaluation of programme to ideal point in the matrix F, its calculating formula is:
Figure 959610DEST_PATH_IMAGE037
，
Figure 395750DEST_PATH_IMAGE031
Wherein,
Figure 393924DEST_PATH_IMAGE038
Be that i is individual by the connect recency of evaluation of programme to ideal point;
Figure DEST_PATH_IMAGE074
Being that i is individual is arrived the most not good enough distance by evaluation of programme,
Figure DEST_PATH_IMAGE076
Be i by the distance of evaluation of programme to ideal point, n is by the number of evaluation of programme, and its result of calculation is:
The described recency that connects
Figure 326511DEST_PATH_IMAGE038
Be one between being value between 0 and 1, to " ideal point "
Figure 44587DEST_PATH_IMAGE023
, its recency that connects
Figure 516458DEST_PATH_IMAGE038
Be 1, to " the most not good enough "
Figure 890457DEST_PATH_IMAGE028
, its recency that connects
Figure 431377DEST_PATH_IMAGE038
Be 0, recency therefore connects
Figure 646370DEST_PATH_IMAGE038
Big more, the approaching more ideal point of point of this scheme is described, good more; Otherwise recency connects
Figure 36945DEST_PATH_IMAGE038
More little, illustrate that the point of this scheme is approaching more the most not good enough, poor more.
(57), according to estimating in the matrix F respectively by the recency that connects of evaluation of programme to ideal point
Figure 757295DEST_PATH_IMAGE038
Each case is made good and bad ordering.
The described recency that connects of step (56)
Figure 167153DEST_PATH_IMAGE038
As can be known, tested bamboo is bamboo 14(Schizostachyum brachycladum most likely), secondly be bamboo 8(Dendrocalamus farinosus) and bamboo 11(Gigantochloa atroviolacea), discriminating is finished.
Method of the present invention is not limited to the embodiment described in the embodiment, and the embodiment of other that those skilled in the art's technical scheme according to the present invention draws belongs to technological innovation scope of the present invention equally.

Claims (6)

1. the plant discrimination method based on cloud model and TOPSIS method is characterized in that this method has at first made up the resemblance sample database of plant; Utilize trapezoidal cloud model to be compared then, obtain the degree of membership compared with resemblance sample database by measuring plants, realized by the preliminary discriminating of measuring plants by the external appearance characteristic sample database of the external appearance characteristic of measuring plants and plant; When identification result when being a plurality of, utilize normal cloud model that result for retrieval is accurately mated calculating again, obtain the accurate degree of membership of being compared with resemblance sample database by measuring plants; Utilize the TOPSIS method that degree of membership is carried out comprehensive evaluation at last, identify plant, concrete steps are as follows:
(1), makes up the resemblance sample database of plant;
(2), utilize trapezoidal cloud model, calculate by the degree of membership of measuring plants, tentatively differentiate by measuring plants;
(3), judge whether to be less than or equal to 1 by the degree of membership of measuring plants, if be to be less than or equal to 1, then change step (5) by the degree of membership of measuring plants, if be to be less than or equal to 1 to equal 1, then change step (4) by the degree of membership of measuring plants;
(4), utilize normal cloud model, be that 1 sample calculates with carried out normal cloud model by measuring plants to degree of membership, obtain by the accurate degree of membership of measuring plants;
(5), utilize the TOPSIS method, degree of membership is carried out comprehensive evaluation identifies by measuring plants.
2. a kind of plant discrimination method according to claim 1 based on cloud model and TOPSIS method, it is characterized in that, the resemblance sample database of structure plant described in the above-mentioned steps (1), these sample data stock is put to can be in order to distinguish the resemblance of floristic plant, and its resemblance value is the numeric type data.
3. a kind of plant discrimination method according to claim 1 based on cloud model and TOPSIS method, it is characterized in that, utilize trapezoidal cloud model described in the above-mentioned steps (2), calculate by the degree of membership of measuring plants, differentiate that tentatively its operation steps is as follows by measuring plants:
(21), determine the expectation interval of trapezoidal cloud model
Figure 585420DEST_PATH_IMAGE001
: according to the resemblance span in the resemblance sample storehouse of plant, determine the expectation interval of trapezoidal cloud model
Figure 72685DEST_PATH_IMAGE001
, wherein
Figure 939184DEST_PATH_IMAGE002
Be the lower limit of this resemblance,
Figure 773845DEST_PATH_IMAGE003
Higher limit for this resemblance value;
(22), calculate trapezoidal cloud model entropy
Figure 595301DEST_PATH_IMAGE004
: utilize the rising cloud and the fall cloud of trapezoidal cloud model, the span of the resemblance of plant is expanded, be set at [3 between the expansion area ,+3
Figure 989953DEST_PATH_IMAGE004
];
(23), the numerical characteristic that utilizes trapezoidal cloud model is to being described by the resemblance of measuring plants: determine trapezoidal cloud expectation curve equation by trapezoidal cloud expectation and entropy:
Figure 221345DEST_PATH_IMAGE005
Figure 264519DEST_PATH_IMAGE006
Wherein,
Figure 124546DEST_PATH_IMAGE007
The degree of membership of ordering for trapezoidal cloud x,
Figure 946396DEST_PATH_IMAGE002
For expecting interval lower limit,
Figure 348690DEST_PATH_IMAGE003
For expecting the interval upper limit,
Figure 121301DEST_PATH_IMAGE004
Entropy for trapezoidal cloud;
(24), utilize trapezoidal cloud model, calculate by the degree of membership of measuring plants: will be utilized trapezoidal cloud model to be analyzed respectively by the resemblance value in all resemblances of measuring plants and the sample storehouse, and obtain by each resemblance of measuring plants
Figure 267243DEST_PATH_IMAGE007
4. a kind of plant discrimination method according to claim 1 based on cloud model and TOPSIS method, it is characterized in that the normal cloud model that utilizes described in the above-mentioned steps (4), degree of membership is 1 sample and is carried out normal cloud model by measuring plants and calculate, obtain by the accurate degree of membership of measuring plants, its operation steps is as follows:
(41), determine the expectation value of normal cloud model:, determine the expectation value of normal cloud model according to profile eigenwert in the resemblance sample storehouse of plant , Intermediate value for the resemblance interval value;
(42), calculate the entropy of normal cloud model
Figure 423308DEST_PATH_IMAGE004
, entropy
Figure 372941DEST_PATH_IMAGE004
Calculating formula be:
(43), can determine normal state cloud expectation curve equation, this curvilinear equation is by normal state cloud expectation and entropy:
Figure 412014DEST_PATH_IMAGE010
Wherein,
Figure 917076DEST_PATH_IMAGE007
The degree of membership of ordering for trapezoidal cloud x,
Figure 404820DEST_PATH_IMAGE011
Be expectation value,
Figure 678938DEST_PATH_IMAGE004
Entropy for trapezoidal cloud;
(44), will to be calculated by the degree of membership of measuring plants by all resemblances of measuring plants and step (2) be that 1 sample is analyzed, and obtains by each resemblance of measuring plants
Figure 859515DEST_PATH_IMAGE007
5. a kind of plant discrimination method based on cloud model and TOPSIS method according to claim 1 is characterized in that utilizing the TOPSIS method described in the above-mentioned steps (5), degree of membership is carried out comprehensive evaluation identify by measuring plants, and its operation steps is as follows:
(51), establish the evaluation matrix F of degree of membership comprehensive evaluation: if the degree of membership value that step (2) is calculated has only one or do not have one to equal at 1 o'clock, the evaluation matrix F is the degree of membership matrix of step (2) gained; Otherwise the evaluation matrix F is that step (3) is calculated the gained result;
(52), determine to estimate the ideal point of matrix F, the calculating formula of its ideal point is:
Figure 586293DEST_PATH_IMAGE012
Wherein,
Figure 877728DEST_PATH_IMAGE013
Be desirable point set,
Figure 271932DEST_PATH_IMAGE014
Be that i is individual by j value of evaluation of programme, Be the objective function numbering collection of asking maximum,
Figure 860922DEST_PATH_IMAGE016
It is the collection of functions of asking minimum target;
(53), determine to estimate the most not good enough of matrix F, its most not good enough calculating formula is:
Figure 690469DEST_PATH_IMAGE017
Wherein,
Figure 689911DEST_PATH_IMAGE018
Be the most not good enough collection,
Figure 949641DEST_PATH_IMAGE019
Be that i is individual by j value of evaluation of programme,
Figure 667324DEST_PATH_IMAGE015
Be the objective function numbering collection of asking maximum,
Figure 552759DEST_PATH_IMAGE016
It is the collection of functions of asking minimum target;
(54), calculate to estimate that each is by the distance of evaluation of programme to ideal point in the matrix F, its calculating formula is:
Figure 635467DEST_PATH_IMAGE020
，
Figure 331678DEST_PATH_IMAGE021
Wherein,
Figure 398640DEST_PATH_IMAGE022
Be that i is individual by the distance of evaluation of programme to ideal point,
Figure 159590DEST_PATH_IMAGE019
Be that i is individual by j value of evaluation of programme,
Figure 887286DEST_PATH_IMAGE023
Be the value of the j item of ideal point, n is by the number of evaluation of programme;
(55), calculate to estimate in the matrix F each by evaluation of programme to the most not good enough distance, its calculating formula is:
Figure 465381DEST_PATH_IMAGE024
，
Figure 144274DEST_PATH_IMAGE021
Wherein,
Figure 997610DEST_PATH_IMAGE025
Being that i is individual is arrived the most not good enough distance by evaluation of programme, Be that i is individual by j value of evaluation of programme,
Figure 890531DEST_PATH_IMAGE026
Be the value of the most not good enough j item, n is by the number of evaluation of programme;
(56), calculate to estimate that each is by the connect recency of evaluation of programme to ideal point in the matrix F, its calculating formula is:
Figure 433551DEST_PATH_IMAGE027
，
Figure 303026DEST_PATH_IMAGE021
Wherein,
Figure 113637DEST_PATH_IMAGE028
Be that i is individual by the connect recency of evaluation of programme to ideal point,
Figure 322552DEST_PATH_IMAGE029
Being that i is individual is arrived the most not good enough distance by evaluation of programme,
Figure 996544DEST_PATH_IMAGE030
Be i by the distance of evaluation of programme to ideal point, n is by the number of evaluation of programme;
(57), according to estimating in the matrix F respectively by the recency that connects of evaluation of programme to ideal point
Figure 313387DEST_PATH_IMAGE028
Each case is made good and bad ordering.
6. a kind of plant discrimination method based on cloud model and TOPSIS method according to claim 5 is characterized in that the trapezoidal cloud model entropy of the described calculating of above-mentioned steps (22)
Figure 100209DEST_PATH_IMAGE004
Its spreading range is 20% of a resemblance lower limit, then trapezoidal cloud model entropy
Figure 227827DEST_PATH_IMAGE004
Calculating formula is: entropy =(
Figure 974295DEST_PATH_IMAGE031
* 0.2)/3.
CN 201110048981 2011-03-02 2011-03-02 Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method Pending CN102156710A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201110048981 CN102156710A (en) 2011-03-02 2011-03-02 Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201110048981 CN102156710A (en) 2011-03-02 2011-03-02 Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method

Publications (1)

Publication Number Publication Date
CN102156710A true CN102156710A (en) 2011-08-17

Family

ID=44438209

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201110048981 Pending CN102156710A (en) 2011-03-02 2011-03-02 Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method

Country Status (1)

Country Link
CN (1) CN102156710A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105740635A (en) * 2016-02-03 2016-07-06 王永林 Cloud ideal solution evaluation method for transformer electromagnetic design scheme
CN109874584A (en) * 2019-03-19 2019-06-14 广州辰轩农业科技有限公司 A kind of fruit tree growing way monitoring system based on deep learning convolutional neural networks
CN112580493A (en) * 2020-12-16 2021-03-30 广东省林业科学研究院 Plant identification method, device and equipment based on unmanned aerial vehicle remote sensing and storage medium
CN113627735A (en) * 2021-07-16 2021-11-09 宁夏建设投资集团有限公司 Early warning method and system for safety risk of engineering construction project

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101344929A (en) * 2008-08-20 2009-01-14 盛秀英 Garden tree classification method
EP2133822A1 (en) * 2008-06-13 2009-12-16 University College Cork A method of stem taper, volume and product breakout prediction
CN101697167A (en) * 2009-10-30 2010-04-21 邱建林 Clustering-decision tree based selection method of fine corn seeds
US20100332475A1 (en) * 2009-06-25 2010-12-30 University Of Tennessee Research Foundation Method and apparatus for predicting object properties and events using similarity-based information retrieval and modeling

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2133822A1 (en) * 2008-06-13 2009-12-16 University College Cork A method of stem taper, volume and product breakout prediction
CN101344929A (en) * 2008-08-20 2009-01-14 盛秀英 Garden tree classification method
US20100332475A1 (en) * 2009-06-25 2010-12-30 University Of Tennessee Research Foundation Method and apparatus for predicting object properties and events using similarity-based information retrieval and modeling
CN101697167A (en) * 2009-10-30 2010-04-21 邱建林 Clustering-decision tree based selection method of fine corn seeds

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105740635A (en) * 2016-02-03 2016-07-06 王永林 Cloud ideal solution evaluation method for transformer electromagnetic design scheme
CN105740635B (en) * 2016-02-03 2018-01-30 中原工学院 A kind of cloud ideal solution evaluation method of transformer electromagnetic design scheme
CN109874584A (en) * 2019-03-19 2019-06-14 广州辰轩农业科技有限公司 A kind of fruit tree growing way monitoring system based on deep learning convolutional neural networks
CN112580493A (en) * 2020-12-16 2021-03-30 广东省林业科学研究院 Plant identification method, device and equipment based on unmanned aerial vehicle remote sensing and storage medium
CN112580493B (en) * 2020-12-16 2021-11-09 广东省林业科学研究院 Plant identification method, device and equipment based on unmanned aerial vehicle remote sensing and storage medium
CN113627735A (en) * 2021-07-16 2021-11-09 宁夏建设投资集团有限公司 Early warning method and system for safety risk of engineering construction project

Similar Documents

Publication Publication Date Title
CN102201001B (en) Fast retrieval method based on inverted technology
CN106845717B (en) Energy efficiency evaluation method based on multi-model fusion strategy
CN101414183B (en) Cigarette working procedure quality overall evaluation system and method based on gray correlation analysis
CN101398836A (en) Search ordering method based on subjectivity and objectivity index and weight allocation
CN102509243A (en) Method and system for evaluating quality in process of manufacturing cigarette
CN103578037A (en) Method and system for evaluating quality of tobacco materials
CN101017476A (en) Characteristic analytical method for product point clouds surface based on dynamic access model
CN102629296A (en) Enterprise credit evaluation method based on gray fuzzy
CN108681744A (en) A kind of electric load curve hierarchy clustering method divided based on data
CN102663447A (en) Cross-media searching method based on discrimination correlation analysis
CN106611016B (en) A kind of image search method based on decomposable word packet model
CN103336771A (en) Data similarity detection method based on sliding window
CN102156710A (en) Plant identification method based on cloud model and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method
CN105574265B (en) Entire assembly model quantitative description towards model index
CN102930291B (en) Automatic K adjacent local search heredity clustering method for graphic image
CN111931992A (en) Power load prediction index selection method and device
CN109284409A (en) Picture group geographic positioning based on extensive streetscape data
CN117668583A (en) Investment optimization method based on artificial intelligent investment research
CN116226468B (en) Service data storage management method based on gridding terminal
CN103049570B (en) Based on the image/video search ordering method of relevant Preserving map and a sorter
CN114676841B (en) Product design evaluation method
Mościbrodzka The Use of Methods of Multidimensional Comparative Analysis in Evaluation of the Standard of Living of Poland’s Population in Comparison with Other Co
CN108182288A (en) A kind of recommendation method based on artificial immunity
CN109254983A (en) A kind of reduction cost method in crowdsourcing TOP-k inquiry
CN110059353A (en) A kind of simplified practical approach of modal parameter automatic identification

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20110817