CN105740918A - Otolith measurement and fish population discriminating method based on polar coordinate - Google Patents

Abstract

The invention relates to an otolith measurement and fish population discriminating method based on a polar coordinate, relates to the fish farming and management field and particularly relates to the method for distinguishing fish populations through measuring fish otolith. According to the method, a center point of an otolith picture is taken as a measurement reference point; end pointd, convex concave points or inflection points having prominent features in the otolith form are searched and taken as characteristic points; through establishing polar coordinates, the characteristic points and a core point are connected, radial length between the core point and each characteristic point is measured to acquire a set of characteristic length data; included angles of two adjacent connection lines are sequentially measured by utilizing the measurement software to acquire a set of characteristic angle data; on the basis of the characteristic length data and the characteristic angle data, a discrimination analysis method or a gray cluster analysis method is utilized, a population type of a detected sample is discriminated. Compared with a discriminating method through utilizing only the length, the otolith measurement and fish population discriminating method is advantaged in that the discrimination result is more accurate and reliable, and the fish population discrimination correct rate can be remarkably improved.

Description

A kind of otolith based on polar angle coordinate is measured and fish population discrimination method

Technical field

The present invention relates to fish culture and management, particularly relate to a kind of method measured by Fish otolith and differentiate the population distinguishing fish.

Background technology

Fishery resources are the important component parts of natural resources, are important sources of human foods, and it is that the people being engaged in fishing activity provide employment, economic interests and social welfare.In many countries, Fish are ingredients important in daily life, be 2/3 world population provide 40% protein, have nearly 1,000,000,000 people to rely on Fish and marine food as their main source of animal protein in Asia.Therefore to fishery resources, especially the hydrobiological research such as Fish just seems most important.

Owing to hydrobiological living environment and mode are comparatively special, understand their growth, age and life cycle, these species are better understood by for us, more reasonably utilize and develop these kinds simultaneously and have very big help, it is increasingly becoming study hotspot in recent years, receives the concern of related researcher.The aquatic creatures live environment such as Fish are water bodys, and this indicates that its life cycle has very big uncertainty, and the biological study such as life habit at researching fish is just different from other animals.Although in the last few years, the labelling direct method for tracing such as release is developed rapidly, various different labels and correlation technique can reflect the living condition of aquatic animal more accurately, but due to different water bodys, to be subject to environmental effect very big, large number of in colony, plus the factor such as natural enemy and material cost, direct research method is also subject to very big restriction.Therefore the indirect method being analyzed the life cycle of these species by other organization material then seems more easy.

Since within 1899, first identify the age of Europe Pleuronectidae with otolith, otolith is always up one of important materials identifying the Fish age so far.The structure that otolith is usually in bony fish calcification at first, not only shows a year deposition, and remains to continued growth when meeting with the unsuitable environmental condition such as hungry, just remains unchanged once be formed, thus becomes the permanent record form of life cycle event.Chinese invention patent application " a kind of method building saury otolith physiological status analytical model " (application number: 201410097976 application publication numbers: CN103843708A) discloses a kind of method building saury otolith physiological status analytical model, including: extract saury biological data, and number after extraction otolith from saury head statocyst and preserve；Utilize micro imaging system CCD that described otolith is taken pictures, take pictures figure in conjunction with microscopic, measure each morphological parameters of saury otolith；Utilize Lycoperdon polymorphum Vitt constellation clustering method that described morphological parameters carries out main constituent morphological characteristic Cluster Analysis, obtain main constituent morphological parameters, obtain otolith speed of growth changing pattern；Otolith is embedded, roughly grinds and polishing；Otolith after polishing is carried out wheel stricture of vagina interpretation and counting, and takes pictures with CCD, judge the age data of otolith according to interpretation and count results；To described age data, and the biological data obtained is analyzed, and draws growth equation.

The research of fish population and colony's problem is a basic work in fishery biology, be the Appropriate application stock of fish first have to solve basic problem.Otolith shape has species specificity highly and significant group is specificity, has important researching value in the fingerling identification and Fish Phylogenetic Studies of Otolith Morphology, and in recent years, many scholars carry out a large amount of basic research with regard to this in the world.Owing to otolith shape has the species specificity of height, being taken as characteristic of division between kind, nearest research concentrates on again it as the value distinguishing Population identification feature.In current analysis method, the feature that morphological analysis is quick with it, simple, accuracy rate is higher, become one of main method of otolith discriminatory analysis.At present, differentiate that population mainly adopts traditional Morphometry according to Otolith Morphology, namely by measuring the radical length of otolith, carry out discriminant analysis in conjunction with ASSOCIATE STATISTICS method, set up discriminant function.But, owing to otolith is less, existing otolith measuring method also exists more error when carrying out linear measure longimetry, eventually has influence on analysis result.

Summary of the invention

It is an object of the invention to provide a kind of otolith based on polar angle coordinate to measure and fish population discrimination method, core point and characteristic point based on Otolith Morphology set up polar coordinate, by otolith radical length and angle angle being combined, make the result that otolith is measured and fish population differentiates more accurate and reliable.

This invention address that above-mentioned technical problem be the technical scheme is that

A kind of otolith based on polar angle coordinate is measured and fish population discrimination method, it is characterised in that comprise the following steps:

S10: first take out otolith in tested sample fish body, clean and be placed under microscope photographing device and take pictures；

S20: using the central point of otolith photo as datum mark, be defined as core point, the form according to otolith, otolith is divided into winged petiole district, back of the body district, back zone and abdomen district；

S30: find out and there is the end points of prominent features, sags and crests or flex point as characteristic point in Otolith Morphology；

S40: by above-mentioned each characteristic point and core point line, with the radical length of Survey Software measurement core point to each characteristic point, it is thus achieved that a stack features length data；

S50: measure the angle between adjacent two lines successively by Survey Software, it is thus achieved that a stack features angular data；

S60: combine described characteristic length data and characteristic angle data, utilize the techniques of discriminant analysis in statistics or grey cluster theory, it determines the kind heap sort of tested sample.

The otolith based on polar angle coordinate according to the present invention measures the one preferably technical scheme with fish population discrimination method, it is characterised in that described step S60 comprises the following steps:

S61: step S40 is measured each characteristic length data obtained and is respectively divided by total length or the fork length of otolith, characteristic length value is standardized；Step S50 is measured the characteristic angle data obtained and is converted into radian value；By the characteristic length data after converting with characteristic angle data combine, obtain m the discrimination variable based on polar angle coordinate；

S62: from G the initial data obtaining N=Nl sample of Σ overall, wherein G is overall number, and Nl is the number of samples of known population l, and wherein, l is general classification sequence number, l=1,2 ... G；

S63: calculate each overall classification average according to following formulaAnd grand mean

${\stackrel{\‾}{\mathrm{Xj}}}^l=\frac{1}{N}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\Σ}}}{X_{\mathrm{aj}}}^l,j=\mathrm{1,2}...m,$

$\stackrel{\‾}{X_J}=\frac{1}{N}\underset{i}{\overset{G}{\mathrm{\Σ}}}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\Σ}}}{X_{\mathrm{aj}}}^l,j=\mathrm{1,2}...m,$

Wherein, m is discrimination variable number；

S64: according to scatter matrix W in following formula calculating group and total scatter matrix T

W=(W_ij)m×m

T=(T_ij)m×m

Wherein $W_{\mathrm{ij}}=\underset{i=1}{\overset{G}{\mathrm{\Σ}}}\underset{i=1}{\overset{\mathrm{Ni}}{\mathrm{\Σ}}}(X_i^l-{\stackrel{\‾}{X_i}}^l)(X_{\mathrm{aj}}^l-X_j^l),t_{\mathrm{ij}}=\underset{i=1}{\overset{G}{\mathrm{\Σ}}}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\Σ}}}(X_{\mathrm{ai}}^l-\stackrel{\‾}{X_i})(X_{\mathrm{aj}}^l-\stackrel{\‾}{X_j}),i,j=\mathrm{1,2}...G;$

S65: arrange the marginal value F1 introducing variable and reject marginal value F2, the usual desirable F1=F2 of variable, adopting stepwise discriminatory method, utilizes F inspection to delete despecking and except variable and introduces variable:

S65a: calculate and each be selected into variable| (L-1), it is assumed that the Ur of r variable | (L-1) that select maximum U, for maximum, does F inspection:If F < F2, then select variable X r as rejecting variable as, it is rejected from discriminant function；

S65b: calculate and be not selected into variableSelect minimum U_i| (L-1), it is assumed that the Ur of r variable | (L), for minimum, does F inspectionIf F > F1, then select variable X r as introducing variable as, be introduced into discriminant function；

S65c: for selected rejecting variable or introducing variable X r, use following formula to eliminate the variable not meeting introducing condition

$W_{\mathrm{ij}}^{(l+1)}=\left\{\begin{array}{ccc}{W}_{\mathrm{ij}}^{\left(l\right)}/W_{\mathrm{rr}}^l& i=r& j\&NotEqual;r\\ W_{\mathrm{ij}}^l-W_{\mathrm{ir}}^l\&CenterDot;W_{\mathrm{rj}}^l/W_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\\ 1/W_{\mathrm{rr}}^l& i=r& j=r\\ -W_{\mathrm{ir}}^l/W_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\end{array}\right\}$

$t_{\mathrm{ij}}^{(l+1)}=\left\{\begin{array}{ccc}{t}_{\mathrm{rj}}^{\left(l\right)}/t_{\mathrm{rr}}^l& i=r& j\&NotEqual;r\\ t_{\mathrm{ij}}^l-t_{\mathrm{ir}}^l\&CenterDot;t_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\\ 1/t_{\mathrm{rr}}^l& i=r& j=r\\ -t_{\mathrm{ir}}^l/t_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\end{array}\right\};$

S66: use the discrimination variable that screening obtains to set up discriminant function and criterion, carries out tested sample differentiating and sorts out:

S66a: according to following formula computational discrimination coefficient

$C_{\mathrm{ig}}=(N-G)\underset{i\&Element;L}{\mathrm{\&Sigma;}}{W}_{\mathrm{ij}}^l\stackrel{\&OverBar;}{X_{\mathrm{ig}}}$

$C_{\mathrm{og}}=\frac{-1}{2}\underset{i\&Element;l}{\mathrm{\&Sigma;}}{C}_{\mathrm{ig}}\stackrel{\&OverBar;}{X_{\mathrm{ig}}}$

S66b: set up discriminant function and criterion $Y_{g\left(x\right)}=\mathrm{Ln}\left(q_g\right)+C_{\mathrm{og}}+\underset{i\&Element;I}{\mathrm{\&Sigma;}}{C}_{\mathrm{ig}}{X}_i,$

If Y_k(x)=max{r_g(x), then x is incorporated into kth class；

Wherein, X_iFor the discrimination variable filtered out, q_gIt it is the overall meansigma methods of g class.

The invention has the beneficial effects as follows:

The otolith based on polar angle coordinate of the present invention is measured and fish population discrimination method, by setting up polar coordinate, determine core point and the characteristic point of Otolith Morphology, radical length and angle angle are combined, it is possible to make the result that otolith is measured and fish population differentiates more accurate and reliable；Compared with the discriminatory analysis method of existing simple use length, the method for the present invention can significantly improve the differentiation accuracy that fish population differentiates.

Accompanying drawing explanation

Fig. 1 is existing Fish otolith measuring method schematic diagram；

Fig. 2 is the present invention based on the otolith of polar angle coordinate measures otolith core point and the characteristic point schematic diagram of process；

Fig. 3 is based on the polar angle coordinate schematic diagram of the Fish otolith measuring method of polar angle coordinate.

Detailed description of the invention

For better understanding of the technique scheme of the present invention, it is described in detail further below in conjunction with drawings and Examples.

The following description adopts the otolith discriminant analysis of sardine and Sardinella aurita as embodiment, sardine and Sardinella aurita sample pick up from 21 ° of 00 ' N～26 ° 00 ' N, marine site, W Morocco exclusive economic zone, 15 ° of 30 ' W～20 ° 00 ', after sampling freezing immediately and take back laboratory, choose 46 tail Sardinella auritas and 53 tail sardine be analyzed.The trunk length of Sardinella aurita ranges for 167-231mm, and sardine trunk length ranges for 185-240mm.

As comparing, it is first according to the existing Fish otolith measuring method shown in Fig. 1, the otolith of each sample is carried out observation under anatomical lens take pictures, the morphological parameters of otolith is measured by Survey Software, six conventional morphological parameters include total length TL (totallength), width TW (totalwidth), carry on the back long DL (dorsallength), across back DW (dorsalwidth), abdomen is laughed long VL (ventrallength) and the long WL of winged petiole (winglength).Measurement result is accurate to 0.01mm, measures and is independently carried out by 2 people, if the error that both measure is more than 5%, then remeasures, otherwise takes their meansigma methods.

The measurement statistical result adopting two kinds of sardine otoliths that existing Fish otolith measuring method obtains is as shown in table 1, when fishery harvesting individual similarly sized, the meansigma methods of the otolith indices of Sardinella aurita is respectively less than sardine.

1 two kinds of sardine otoliths of table are measured parameter and are compared (traditional measurement)

Then, measure and fish population discrimination method according to the otolith based on polar angle coordinate of the present invention, carry out otolith measurement according to following steps:

S10: first take out otolith in tested sample fish body, clean and be placed under microscope photographing device and take pictures；Photo is it is ensured that otolith clear-cut.

S20: using the central point of otolith photo as datum mark, is defined as core point O, the form according to otolith, and otolith is divided into winged petiole district, back of the body district, back zone and abdomen district；

S30: find out and there is the end points of prominent features, sags and crests or flex point as characteristic point in Otolith Morphology, described characteristic point is also referred to as fixed point, including winged petiole terminal A, the back of the body summit, district B, back porch intersection point C, summit, back zone D, Hou Fu district intersection point E, abdomen district summit F, abdomen district far point G and Yi Fu district intersection point H, referring to Fig. 2；

S40: by above-mentioned each characteristic point and core point O line, with the radical length of Survey Software WT-3000G measurement core point O to each characteristic point, it is thus achieved that a stack features length data, the embodiment according to Fig. 2 and Fig. 3, obtain 8 characteristic length data, be designated as OA, OB, OC respectively, OD, OE, OF, OG and OH；

Table 2 is that the two kinds of sardine otoliths adopting the method measurement of the present invention to obtain measure parameter result simultaneously, can draw from table 2, and except OA, OG and OH, remaining length value Sardinella aurita is respectively less than sardine, different with the length value of traditional measurement.

2 two kinds of sardine otoliths of table are measured parameter and are compared (polar angle measurement of coordinates)

S50: measure the angle between adjacent two lines successively with Survey Software WT-3000G, it is thus achieved that a stack features angular data, the embodiment according to Fig. 2 and Fig. 3, obtain 8 characteristic angle data, be designated as a, b, c, d, e, f, g and h respectively；

Can drawing from table 3, except c, f and g, remaining angle value Sardinella aurita is all higher than sardine, and the angle of otolith each several part is different.

3 two kinds of sardine otolith angle parameters of table compare (polar angle measurement of coordinates)

S60: combine described characteristic length data and characteristic angle data, utilize the techniques of discriminant analysis in statistics or grey cluster theory, it determines the kind heap sort of tested sample.

The otolith based on polar angle coordinate according to the present invention measures an embodiment with fish population discrimination method, and described step S60 comprises the following steps:

S61: step S40 is measured the long FL of the fork (forklength) that each characteristic length data obtained are respectively divided by otolith, characteristic length value is standardized, eliminates the allometric impact of tested sample；Step S50 is measured the characteristic angle data obtained and is converted into radian value；By the characteristic length data after converting with characteristic angle data combine, obtain m the discrimination variable based on polar angle coordinate, in the embodiment of Fig. 2 and Fig. 3, m=8, namely angle and length have 8 variablees respectively；Characteristic length value standardization can also be respectively divided by total length TL by measuring each characteristic length data obtained.

S62: from G the initial data obtaining N=Nl sample of Σ overall, wherein G is overall number, and Nl is the number of samples of known population l, and wherein, l is general classification sequence number, l=1,2 ... G；

S63: calculate each overall classification average according to following formulaAnd grand mean

${\stackrel{\&OverBar;}{\mathrm{Xj}}}^l=\frac{1}{N}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}{X_{\mathrm{aj}}}^l,j=\mathrm{1,2}...m,$

$\stackrel{\&OverBar;}{X_J}=\frac{1}{N}\underset{i}{\overset{G}{\mathrm{\&Sigma;}}}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}{X_{\mathrm{aj}}}^l,j=\mathrm{1,2}...m,$

Wherein, m is discrimination variable number, m=8 in the embodiment of Fig. 2 and Fig. 3；

S64: according to scatter matrix W in following formula calculating group and total scatter matrix T

W=(W_ij)m×m

T=(T_ij)m×m

Wherein $W_{\mathrm{ij}}=\underset{i=1}{\overset{G}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}(X_i^l-{\stackrel{\&OverBar;}{X_i}}^l)(X_{\mathrm{aj}}^l-X_j^l),t_{\mathrm{ij}}=\underset{i=1}{\overset{G}{\mathrm{\&Sigma;}}}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}(X_{\mathrm{ai}}^l-\stackrel{\&OverBar;}{X_i})(X_{\mathrm{aj}}^l-\stackrel{\&OverBar;}{X_j}),i,j=\mathrm{1,2}...G;$

S65: arrange the marginal value F1 introducing variable and reject marginal value F2, the usual desirable F1=F2 of variable, adopting stepwise discriminatory method, utilizes F inspection to delete despecking and except variable and introduces variable:

S65a: calculate and each be selected into variable| (L-1), it is assumed that the Ur of r variable | (L-1) that select maximum U, for maximum, does F inspection:If F < F2, then select variable X r as rejecting variable as, it is rejected from discriminant function；

S65b: calculate and be not selected into variableSelect minimum U_i| (L-1), it is assumed that the Ur of r variable | (L), for minimum, does F inspectionIf F > F1, then select variable X r as introducing variable as, be introduced into discriminant function；

S65c: for selected rejecting variable or introducing variable X r, use following formula to eliminate the variable not meeting introducing condition

$W_{\mathrm{ij}}^{(l+1)}=\left\{\begin{array}{ccc}{W}_{\mathrm{ij}}^{\left(l\right)}/W_{\mathrm{rr}}^l& i=r& j\&NotEqual;r\\ W_{\mathrm{ij}}^l-W_{\mathrm{ir}}^l\&CenterDot;W_{\mathrm{rj}}^l/W_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\\ 1/W_{\mathrm{rr}}^l& i=r& j=r\\ -W_{\mathrm{ir}}^l/W_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\end{array}\right\}$

$t_{\mathrm{ij}}^{(l+1)}=\left\{\begin{array}{ccc}{t}_{\mathrm{rj}}^{\left(l\right)}/t_{\mathrm{rr}}^l& i=r& j\&NotEqual;r\\ t_{\mathrm{ij}}^l-t_{\mathrm{ir}}^l\&CenterDot;t_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\\ 1/t_{\mathrm{rr}}^l& i=r& j=r\\ -t_{\mathrm{ir}}^l/t_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\end{array}\right\};$

S66: use the discrimination variable that screening obtains to set up discriminant function and criterion, carries out tested sample differentiating and sorts out:

S66a: according to following formula computational discrimination coefficient

$C_{\mathrm{ig}}=(N-G)\underset{i\&Element;L}{\mathrm{\&Sigma;}}{W}_{\mathrm{ij}}^l\stackrel{\&OverBar;}{X_{\mathrm{ig}}}$

$C_{\mathrm{og}}=\frac{-1}{2}\underset{i\&Element;l}{\mathrm{\&Sigma;}}{C}_{\mathrm{ig}}\stackrel{\&OverBar;}{X_{\mathrm{ig}}}$

S66b: set up discriminant function and criterion

$Y_{g\left(x\right)}=\mathrm{Ln}\left(q_g\right)+C_{\mathrm{og}}+\underset{i\&Element;I}{\mathrm{\&Sigma;}}{C}_{\mathrm{ig}}{X}_i,$

If Y_k(x)=max{r_g(x), then x is incorporated into kth class；

Wherein, X_iFor the discrimination variable filtered out, q_gIt it is the overall meansigma methods of g class.

In order to the technique effect of the present invention is described, existing by as follows for the effectiveness comparison using Stepwise Discriminatory Analysis that two kinds of sardine are carried out discriminant analysis:

The measurement method using prior art is adopted to obtain 6 measured value TL, TW, DL, DW, VL and WL, adopts Wilks ' Lambda method that 6 parameter values are screened, and have finally chosen across back (DW/FL), the back of the body long (DL/FL) and 3 parameters of width (TW/FL), two kinds of sardine are classified, and the result of its variance analysis all shows that three indexs have notable contribution (P < 0.01).Adopting this three Index Establishment discriminant functions, concrete formula is as follows:

Sardinella aurita: Y=14.384 × DW/FL+13.663 × DL/FL+0.4603 × TW/FL-0.145,

Sardine: Y=16.414 × DW/FL+12.376 × DL/FL+3.0162 × TW/FL-0.154.

The Otolith Morphology index picked out by two kinds of sardine samples is brought in above-mentioned discriminant function, then this sample is included into the bigger function of gained Y value is its corresponding classification.Two kinds of sardine discriminant analysis results based on existing measurement method are as shown in table 4, and from differentiating result, Sardinella aurita differentiates that accuracy is 84.78%, and sardine is 88.68%, and overall accuracy is 86.73%.With initial, the result of validation-cross differentiates that result is similar, Sardinella aurita and sardine respectively 82.61% and 84.92%, overall accuracy is 83.76%.

Table 4 is based on two kinds of sardine discriminant analysis results of traditional measurement

The otolith based on polar angle coordinate adopting the present invention is measured and fish population discrimination method, measure 8 the line segment length OA obtained, OB, OC, OD, OE, OF, OG and OH, and 8 angle value a, b, c, d, e, f, g and h, then length value is standardized, angle value is converted to Circular measure, it is analyzed collectively as variable, adopt Wilks ' Lambda method that 16 parameter values are screened, have finally chosen 5 length value (OB/FL, OC/FL, OF/FL, OG/FL and OH/FL) and 2 angle values (a and g), totally 7 parameters are as discrimination variable, two kinds of sardine are classified, the result of its variance analysis all shows that three indexs have notable contribution (P < 0.01).Adopting this three Index Establishment discriminant functions, concrete formula is as follows:

Sardinella aurita: Y=15.455 × OB/FL+59.180 × OC/FL-139.583 × OF/FL+205.527 × OG/FL-91.694 × OH/FL+297.664 × a+259.399 × g-388.772

Sardine: Y=68.066 × OB/FL+24.106 × OC/FL-87.101 × OF/FL+184.362 × OG/FL-121.359 × OH/FL+311.228 × a+278.782 × g-404.514

Table 5 is based on two kinds of sardine discriminant analysis results of polar angle coordinate survey

The Otolith Morphology index picked out by two kinds of sardine samples is brought in above-mentioned discriminant function, then this sample is included into the bigger function of gained Y value is its corresponding classification.From differentiating result, Sardinella aurita differentiates that accuracy is 93.54%, and sardine is 98.12%, and overall accuracy is 95.83%.With initial, the result of validation-cross differentiates that result is similar, Sardinella aurita and sardine respectively 89.17% and 98.12%, overall accuracy is 93.65%, referring to table 5.It can be seen that adopt the otolith based on polar angle coordinate of the present invention to measure and fish population discrimination method, it determines accuracy is apparently higher than the measurement method of discrimination of prior art.

Those of ordinary skill in the art will be appreciated that; above embodiments is intended merely to explanation technical scheme; and it is not used as limitation of the invention; any embodiment described above is made based on the connotation of the present invention change, modification, all by the scope of the claims dropping on the present invention.

Claims (2)

1. the otolith based on polar angle coordinate is measured and fish population discrimination method, it is characterised in that comprise the following steps:

S10: first take out otolith in tested sample fish body, clean and be placed under microscope photographing device and take pictures；

S20: using the central point of otolith photo as datum mark, be defined as core point, the form according to otolith, otolith is divided into winged petiole district, back of the body district, back zone and abdomen district；

S30: find out and there is the end points of prominent features, sags and crests or flex point as characteristic point in Otolith Morphology；

S40: by above-mentioned each characteristic point and core point line, with the radical length of Survey Software measurement core point to each characteristic point, it is thus achieved that a stack features length data；

S50: measure the angle between adjacent two lines successively by Survey Software, it is thus achieved that a stack features angular data；

S60: combine described characteristic length data and characteristic angle data, utilize the techniques of discriminant analysis in statistics or grey cluster theory, it determines the kind heap sort of tested sample.

2. the otolith based on polar angle coordinate according to claim 1 is measured and fish population discrimination method, it is characterised in that described step S60 comprises the following steps:

S61: step S40 is measured each characteristic length data obtained and is respectively divided by total length or the fork length of otolith, characteristic length value is standardized；Step S50 is measured the characteristic angle data obtained and is converted into radian value；By the characteristic length data after converting with characteristic angle data combine, obtain m the discrimination variable based on polar angle coordinate；

S62: from G the initial data obtaining N=Nl sample of Σ overall, wherein G is overall number, and Nl is the number of samples of known population l, and wherein, l is general classification sequence number, l=1,2 ... G；

S63: calculate each overall classification average according to following formulaAnd grand mean

${\stackrel{\&OverBar;}{\mathrm{Xj}}}^l=\frac{1}{N}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}{X_{\mathrm{aj}}}^l,j=\mathrm{1,2}...m,$

$\stackrel{\&OverBar;}{X_J}=\frac{1}{N}\underset{i}{\overset{G}{\mathrm{\&Sigma;}}}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}{\mathrm{Xaj}}^l,j=\mathrm{1,2}...m,$

Wherein, m is discrimination variable number；

S64: according to scatter matrix W in following formula calculating group and total scatter matrix T

W=(W_ij)m×m

T=(T_ij)m×m

Wherein $W_{\mathrm{ij}}=\underset{i=1}{\overset{G}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}(X_i^l-{\stackrel{\&OverBar;}{X}}_i^l)(X_{\mathrm{aj}}^l-X_j^l),t_{\mathrm{ij}}=\underset{i=1}{\overset{G}{\mathrm{\&Sigma;}}}\underset{a=1}{\overset{\mathrm{Ni}}{\mathrm{\&Sigma;}}}(X_{\mathrm{ai}}^l-\stackrel{\&OverBar;}{X_i})(X_{\mathrm{aj}}^l-\stackrel{\&OverBar;}{X_j}),$ I, j=1,2...G；

S65: arrange the marginal value F1 introducing variable and reject marginal value F2, the usual desirable F1=F2 of variable, adopting stepwise discriminatory method, utilizes F inspection to delete despecking and except variable and introduces variable:

S65a: calculate and each be selected into variable| (L-1), it is assumed that the Ur of r variable | (L-1) that select maximum U, for maximum, does F inspection:If F < F2, then select variable X r as rejecting variable as, it is rejected from discriminant function；

S65b: calculate and be not selected into variableSelect minimum U_i| (L-1), it is assumed that the Ur of r variable | (L), for minimum, does F inspectionIf F > F1, then select variable X r as introducing variable as, be introduced into discriminant function；

S65c: for selected rejecting variable or introducing variable X r, use following formula to eliminate the variable not meeting introducing condition

$W_{\mathrm{ij}}^{\left(\text{l+1}\right)}=\left\{\begin{array}{ccc}{W}_{\mathrm{rj}}^{\left(l\right)}/W_{\mathrm{rr}}^l& i=r& j\&NotEqual;r\\ W_{\mathrm{ij}}^l-W_{\mathrm{ir}}^l\&CenterDot;W_{\mathrm{rj}}^l/W_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\\ 1/W_{\mathrm{rr}}^l& i=r& j=r\\ -W_{\mathrm{ir}}^l/W_{\mathrm{rr}}^l& i\&NotEqual;\text{r}& j\&NotEqual;r\end{array}\right\}$

$t_{\mathrm{ij}}^{\left(\text{l+1}\right)}=\left\{\begin{array}{ccc}{t}_{\mathrm{rj}}^{\left(l\right)}/t_{\mathrm{rr}}^l& i=r& j\&NotEqual;r\\ t_{\mathrm{ij}}^l-t_{\mathrm{ir}}^l\&CenterDot;t_{\mathrm{rr}}^l& i\&NotEqual;r& j\&NotEqual;r\\ 1/t_{\mathrm{rr}}^l& i=r& j=r\\ -t_{\mathrm{ir}}^l/t_{\mathrm{rr}}^l& i\&NotEqual;\text{r}& j\&NotEqual;r\end{array}\right\};$

S66: use the discrimination variable that screening obtains to set up discriminant function and criterion, carries out tested sample differentiating and sorts out:

S66a: according to following formula computational discrimination coefficient

$C_{\mathrm{ig}}=(N-G)\underset{i\&Element;L}{\mathrm{\&Sigma;}}{W}_{\mathrm{ij}}^l\stackrel{\&OverBar;}{X_{\mathrm{ig}}}$

$C_{\mathrm{og}}=\frac{-1}{2}\underset{i\&Element;1}{\mathrm{\&Sigma;}}{C}_{\mathrm{ig}}\stackrel{\&OverBar;}{X_{\mathrm{ig}}}$

S66b: set up discriminant function and criterion

$Y_{g\left(x\right)}=\mathrm{Ln}\left(q_g\right)+C_{\mathrm{og}}+\underset{i\&Element;I}{\mathrm{\&Sigma;}}{C}_{\mathrm{ig}}{X}_i,$

If Y_k(x)=max{r_g(x), then x is incorporated into kth class；

Wherein, Xi is the discrimination variable filtered out, q_gIt it is the overall meansigma methods of g class.