CN113486940A - Migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth - Google Patents

Migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth Download PDF

Info

Publication number
CN113486940A
CN113486940A CN202110737780.8A CN202110737780A CN113486940A CN 113486940 A CN113486940 A CN 113486940A CN 202110737780 A CN202110737780 A CN 202110737780A CN 113486940 A CN113486940 A CN 113486940A
Authority
CN
China
Prior art keywords
observation
bird
lon
lat
data
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
Application number
CN202110737780.8A
Other languages
Chinese (zh)
Other versions
CN113486940B (en
Inventor
杨秦敏
冯时
乔慧捷
陈积明
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.)
Zhejiang University ZJU
Original Assignee
Zhejiang University ZJU
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 Zhejiang University ZJU filed Critical Zhejiang University ZJU
Priority to CN202110737780.8A priority Critical patent/CN113486940B/en
Publication of CN113486940A publication Critical patent/CN113486940A/en
Application granted granted Critical
Publication of CN113486940B publication Critical patent/CN113486940B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—2D [Two Dimensional] image generation
    • G06T11/20—Drawing from basic elements, e.g. lines or circles
    • G06T11/203—Drawing of straight lines or curves

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth. Some migratory birds cannot implement individual tracking means due to their extremely small body sizes. In order to establish an effective research protection mechanism for the migratory birds, the invention plans to reconstruct the population track in the bird full migration period by using the bird watching information list according to an eBird citizen data platform. Firstly, filling missing values in bird watching information by linear interpolation; then cleaning and removing the abnormal data points; on the basis, clustering is carried out on the bird watching data every day by using an improved Mean-shift clustering algorithm of self-adaptive bandwidth, clustering judgment is carried out on the clustering results every day by using a traversal algorithm based on minimum cost, and nonlinear fitting is realized by using a generalized additive model. The method has important scientific significance and application value for the track reconstruction of the small migratory birds which lack accurate tracking data in the migration period, and further provides scientific support for targeted migratory bird protection activities and protected area construction.

Description

Migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth
Technical Field
The invention relates to a migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth, which is used for automatically grouping and multi-trajectory reconstruction of mass observation data of citizens and belongs to the field of ecology.
Background
The current rapid global environment and climate change poses potential challenges for migrating species that need to cope with or adapt to new conditions. Some migratory birds are particularly sensitive to environmental and climatic changes due to their extremely small size, high metabolic rate and dependence on nectar, grass seeds and the like as main resources, and individual tracking means such as satellites or GPS cannot be implemented due to the limitation of their own volume and weight. In order to establish an effective research and protection mechanism for the migratory birds and perform tracking protection in the whole migration period, massive migratory bird observation information from an eBird national science database is used for reconstructing a migration track of the migratory birds in the annual period so as to explore important information such as population change conditions, speed change conditions and stopover places of the migratory birds in the migration period. According to an eBird citizen data platform created by a Kannell bird laboratory, the invention is supposed to carry out population multi-track reconstruction in a bird full migration period by using bird names, longitudes, latitudes, bird watching time and other information in a bird watching information list.
The Mean-Shift clustering algorithm is an unsupervised clustering algorithm based on density, achieves the purpose of clustering by continuously iterating and searching for the region with the highest density, does not need to make the clustering number in advance, is an unsupervised clustering algorithm with wide application, has fewer used limiting conditions and has strong universality.
Disclosure of Invention
The invention aims to perform multi-track reconstruction in the migration annual cycle of the migratory birds by using mass citizen bird watching data aiming at small migratory birds which cannot be accurately tracked, so that dynamic information in the migration process can be mined, and effective suggestions for setting a protected area can be provided.
The purpose of the invention is realized by the following technical scheme: a migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth comprises the following steps:
(1) acquiring observation information LAT, LON and D of a certain migrant bird in a year period on a citizen data platform eBird, wherein the acquired observation information is stored in an array form and is expressed as follows:
Figure BDA0003142180490000021
wherein, LATjThe latitude, LON of the jth observation information of the migratory birdjLongitude of j-th observation of the waiting bird, DjIndicating the j day, n is the number of acquired bird-waiting observation information, m is the number of days of acquired bird-waiting observation information, and m belongs to [1,366 ]],n≥m；
(2) If the repeated observation information data exists, deleting the same observation longitude and latitude information on the same date to reduce the redundancy of the data set, wherein the data is represented as follows: when any observation date D is selectedjWhen the temperature of the water is higher than the set temperature,
LATp≠LATqand LONp≠LONq
Wherein, LATp,LATqLatitude, LON of any two observation information of the j-th birdp,LONqLongitude of any two pieces of observation information of the j-th bird;
(3) converting longitude and latitude coordinates LAT and LON of the observed geographic position of the migratory bird into mercator coordinates M-LAT and M-LON, and recording as:
Figure BDA0003142180490000022
(4) to ensure the time continuity of the data, the observation date D missing in the original data is usedsAnd its corresponding observed geographic information LATs,LONsK nearest neighbor samples (generally, k takes 2) are selected, the euclidean distance is taken as a distance function, and linear interpolation is performed to obtain:
Figure BDA0003142180490000023
Figure BDA0003142180490000024
Figure BDA0003142180490000025
wherein D ist,DlIs DsObservation date of k-neighbors, LATt,LATlIs LATsLatitude information of observation site of k-nearest neighbor, LONt,LONlIs LONsk-observation place longitude information of adjacent neighbors, wherein l is less than s and less than t;
merging the obtained interpolation into the data set without interpolation to obtain:
Figure BDA0003142180490000031
wherein, N is the number of the bird observation information obtained after interpolation, M is the number of days of the bird observation information obtained after interpolation, and M is 365 or 366; LM-LAT, LM-LON and LD are observation information longitude and latitude and observation date obtained after interpolation;
(5) in order to eliminate abnormal data points existing in the citizen observation data set, aiming at the characteristics of the space data set, an abnormal point detection algorithm based on space data point influence factors is adopted:
define LM-LAT, LM-LON as data set X ═ X1,X2,...XN}, set of weights
Figure BDA0003142180490000032
The spatial k distance neighborhood N (k, O) of the spatial object O can be represented as:
Figure BDA0003142180490000033
where k is the distance of a given spatial object o and dist (p, o, ω) is the distance of spatial objects p and o;
spatial object XiAnd XjThe distance of (d) can be defined as:
Figure BDA0003142180490000034
wherein x isik,xjkAre each Xi,XjF is a normalization function;
the average μ of the object to all neighbors in its neighborhood can be written as:
Figure BDA0003142180490000035
wherein N isK(p) space k distance neighborhood, N, representing object pK(p) | is the number of all neighbors in the neighborhood;
the spatial offset rate of the object p
Figure BDA0003142180490000036
Comprises the following steps:
Figure BDA0003142180490000041
spatial offset impact Rate of object p
Figure BDA0003142180490000042
Comprises the following steps:
Figure BDA0003142180490000043
the spatial offset impact factor of object p can be derived
Figure BDA0003142180490000044
Comprises the following steps:
Figure BDA0003142180490000045
and removing abnormal points with larger deviation influence factors in the observed data points to improve the data quality, and obtaining:
Figure BDA0003142180490000046
wherein r is the number of the observation information of the migratory birds obtained after the abnormal points are removed, and DLM-LAT and DLM-LON are the longitude and latitude of the observation information obtained by processing through an abnormal point detection algorithm;
(6) because the difference of the daily observation data amount of a certain migratory bird in the observation data set is large, the observation data of continuous seven days are aggregated into the observation data of the first day in the seven days by adopting a sliding window idea to increase the stability of clustering, and the method comprises the following steps:
Figure BDA0003142180490000047
the HDLM-LAT and the HDLM-LON are longitude and latitude of observation information obtained through sliding window thought aggregation;
(7) clustering and clustering daily observing data of the migratory birds by adopting a Mean-Shift algorithm based on self-adaptive bandwidth, and generating the size of a Mean-Shift window by utilizing a bandwidth estimation algorithm, wherein the size is expressed by randomly selecting P samples from the observing data of the migratory birds to be classified, and for any sample point PiDefining the distance of each pair of samples in the unsupervised n neighbor (the ratio of neighbor to sample in neighbor search is suggested to be 0.3-0.7) search result as follows:
M＝(m1,m2,...,mn)
selecting the maximum distance in M
Figure BDA0003142180490000051
Calculating the sum of the farthest pairing distances of the unsupervised n-neighbor search of the P samples, and recording as:
Figure BDA0003142180490000052
then selecting
Figure BDA0003142180490000053
As the final bandwidth, obtaining an unsupervised observation clustering result every day by using a Mean-Shift algorithm;
(8) according to the unsupervised observation clustering condition of each day, performing distance traversal on the clustering centers of two adjacent days by using a traversal algorithm based on minimum cost:
calculating the distance D between any two clustering centers in two adjacent days by using a tree structure hierarchical traversal method:
Figure BDA0003142180490000054
wherein (x)1,x2),(y1,y2) D is a coordinate under the mercator coordinate system of the clustering center and is in direct proportion to the cost;
obtaining the grouping condition of the population clustering center in the annual cycle;
Figure BDA0003142180490000055
wherein T represents the total group number of the migratory bird groups, which is equal to the number of the finally fitted tracks;
(9) utilizing a Generalized Additive Model (GAM) to carry out nonlinear fitting on longitude and latitude of the migratory bird observation packet data in an annual period and time respectively so as to obtain the functional relation between longitude latitude and latitude longtude and time t:
latitude＝g(t)；longitude＝h(t)
(10) and finding out the corresponding longitude and latitude coordinates on the same date, and connecting the geographic positions of all the dates to obtain the final migratory bird migration track reconstruction result.
The invention has the beneficial effects that: in order to establish an effective research protection mechanism for the migratory birds, the invention plans to reconstruct the population track in the bird full migration period by using the bird watching information list according to an eBird citizen data platform. Firstly, linear interpolation filling is carried out on missing values in bird watching information, and time continuity of bird watching data is kept; then, an abnormal data point detection algorithm based on space data point influence factors is adopted to clean and remove the abnormal data points, so that the data quality is improved; on the basis, density clustering is carried out on bird watching data every day by using an improved unsupervised clustering algorithm Mean-shift of self-adaptive bandwidth, clustering judgment is carried out on clustering results every day by using a traversal algorithm based on minimum cost, nonlinear fitting is realized by using a generalized additive model, and a reconstructed multi-track result is obtained. The method has important scientific significance and application value for the track reconstruction of the small migratory birds which lack accurate tracking data in the migration period, and further provides scientific support for targeted migratory bird protection activities and protected area construction.
Drawings
FIG. 1 is a flow chart of a migratory bird migration trajectory reconstruction method based on adaptive bandwidth;
FIG. 2 is a graph showing a comparison of results before and after the detection of outliers in the example of the present invention;
FIG. 3 is a diagram illustrating hierarchical traversal clustering, according to an embodiment of the present invention;
FIGS. 4 (a) and (b) are schematic diagrams of longitude and latitude as a function of time, respectively, in an embodiment of the present invention;
FIG. 5 is a schematic diagram of 4 migratory bird migration trajectories fitted by an embodiment of the present invention;
FIG. 6 is a comparison graph of the reconstructed trajectory and the distribution of the original observed data in the embodiment of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings and specific embodiments. The following examples are only for illustrating the technical solutions of the present invention more clearly, and should not be taken as limiting the scope of the present invention.
As shown in fig. 1, a method for reconstructing a migratory bird migration trajectory based on adaptive bandwidth provided in an embodiment of the present invention includes the following steps:
(1) in this example, taking an Anthus _ Spragueii in North America as an example, the 2018 observation information arrays LAT, LON, D on the citizen data platform eBird are obtained and expressed as:
Figure BDA0003142180490000061
wherein, LATjThe latitude, LON of the jth observation information of the migratory birdjLongitude of j-th observation of the waiting bird, DjIndicating day j.
(2) The same observed latitude and longitude information on the same date is checked and deleted to reduce data set redundancy, which is expressed as: when any observation date D is selectedjWhen the temperature of the water is higher than the set temperature,
LATp≠LATqand LONp≠LONq
Wherein, LATp,LATqLatitude, LON of any two observation information of the j-th birdp,LONqThe longitude of any two pieces of observation information of the j-th bird.
(3) Converting longitude and latitude coordinates LAT and LON of the observed geographic position of the migratory bird into mercator coordinates M-LAT and M-LON, and recording as:
Figure BDA0003142180490000071
(4) to ensure the time continuity of the data, the observation date D missing in the original data is usedsAnd its corresponding observed geographic information LATs,LONsK nearest neighbor samples (generally, k takes 2) are selected, the euclidean distance is taken as a distance function, and linear interpolation is performed to obtain:
Figure BDA0003142180490000072
Figure BDA0003142180490000073
Figure BDA0003142180490000074
wherein D ist,DlIs DsObservation date of k-neighbors, LATt,LATlIs LATsLatitude information of observation site of k-nearest neighbor, LONt,LONlIs LONsk-observation place longitude information of adjacent neighbors, wherein l is less than s and less than t;
merging the obtained interpolation into the data set without interpolation to obtain:
Figure BDA0003142180490000075
wherein, LM-LAT, LM-LON and LD are observation information longitude and latitude and observation date obtained by interpolation.
(5) In order to eliminate abnormal data points existing in the citizen observation data set, aiming at the characteristics of the space data set, an abnormal point detection algorithm based on space data point influence factors is adopted:
define LM-LAT, LM-LON as data set X ═ X1,X2,...XN}, set of weights
Figure BDA0003142180490000081
The spatial k distance neighborhood N (k, O) of the spatial object O can be represented as:
Figure BDA0003142180490000082
where k is the distance of a given spatial object o and dist (p, o, ω) is the distance of spatial objects p and o;
spatial object XiAnd XjThe distance of (d) can be defined as:
Figure BDA0003142180490000083
wherein x isik,xjkAre respectively provided withIs Xi,XjF is a normalization function;
the average μ of the object to all neighbors in its neighborhood can be written as:
Figure BDA0003142180490000084
wherein N isK(p) space k distance neighborhood, N, representing object pK(p) | is the number of all neighbors in the neighborhood;
the spatial offset rate of the object p
Figure BDA0003142180490000085
Comprises the following steps:
Figure BDA0003142180490000086
spatial offset impact Rate of object p
Figure BDA0003142180490000087
Comprises the following steps:
Figure BDA0003142180490000088
the spatial offset impact factor of object p can be derived
Figure BDA0003142180490000089
Comprises the following steps:
Figure BDA00031421804900000810
and removing abnormal points with larger deviation influence factors in the observed data points to improve the data quality, and obtaining:
Figure BDA00031421804900000811
wherein DLM-LAT and DLM-LON are observation information longitude and latitude obtained through processing of an abnormal point detection algorithm;
FIG. 2 is a graph showing a comparison of the results before and after the detection of the singular point in this example.
(6) Because the difference of the daily observation data amount of a certain migratory bird in the observation data set is large, the observation data of continuous seven days are aggregated into the observation data of the first day in the seven days by adopting a sliding window idea to increase the stability of clustering, and the method comprises the following steps:
Figure BDA0003142180490000091
the HDLM-LAT and the HDLM-LON are longitude and latitude of observation information obtained through sliding window thought aggregation.
(7) Clustering and clustering daily observing migratory bird data by adopting a Mean-Shift algorithm based on self-adaptive bandwidth, and generating the size of a Mean-Shift window by utilizing a bandwidth estimation algorithm, wherein the size is represented by randomly selecting P samples from the observing migratory bird data to be classified, in the embodiment, P is 500, and for any sample point PiDefining the distance of each pair of samples in the unsupervised n-neighbor (the ratio of neighbor in neighbor search to sample is 0.5 in this embodiment) search result as:
M＝(m1,m2,...,mn)
selecting the maximum distance in M
Figure BDA0003142180490000092
Calculating the maximum pairing distance sum of 500 sample unsupervised n-neighbor search, and recording the sum
Figure BDA0003142180490000093
Then selecting
Figure BDA0003142180490000094
As final bandwidth, a Mean-Shift algorithm was used to obtain daily unsupervised observed clustering results.
(8) According to the unsupervised observation clustering condition of each day, performing distance traversal on the clustering centers of two adjacent days by using a traversal algorithm based on minimum cost:
calculating the distance D between any two clustering centers in two adjacent days by using a tree structure hierarchical traversal method:
Figure BDA0003142180490000095
wherein (x)1,x2),(y1,y2) D is a coordinate under the mercator coordinate system of the clustering center and is in direct proportion to the cost;
obtaining the grouping condition of the population clustering center in the annual cycle:
Figure BDA0003142180490000101
Figure BDA0003142180490000102
in this embodiment, the number of the migratory bird groups is 4, which is equal to the number of the finally fitted tracks;
fig. 3 is a schematic diagram of performing rank traversal in this embodiment, two layers of migratory icons respectively represent Mean-Shift unsupervised clustering results of two adjacent days, the migratory cluster centers of labels 1,2, and 3 respectively perform distance traversal calculation with the migratory cluster centers of labels 4 and 5, and for the migratory cluster center of label 1, L (1,4) is the minimum distance after traversal calculation, so that two populations represented by the migratory cluster centers of labels 1 and 4 can be considered as the same population, and so on, and clustering operation is completed.
(9) Utilizing a Generalized Additive Model (GAM) to carry out nonlinear fitting on longitude and latitude of the migratory bird observation packet data in an annual period and time respectively so as to obtain the functional relation between longitude latitude and latitude longtude and time t:
latitude＝g(t)；longitude＝h(t)
fig. 4 (a) and (b) are schematic diagrams of longitude and latitude as a function of time in the embodiment of the present invention, respectively.
(10) And finding out the corresponding longitude and latitude coordinates on the same date, and connecting the geographic positions of all the dates to obtain the final migratory bird migration track reconstruction result.
FIG. 5 is a schematic diagram of 4 migratory bird migration trajectories fitted by an embodiment of the present invention; FIG. 6 is a schematic diagram of a comparison of a reconstructed trajectory with an original observed data distribution in an embodiment of the present invention; therefore, the method can obtain a better reconstruction effect of the migratory bird migration track.
The foregoing is only a preferred embodiment of the present invention, and although the present invention has been disclosed in the preferred embodiments, it is not intended to limit the present invention. Those skilled in the art can make numerous possible variations and modifications to the present teachings, or modify equivalent embodiments to equivalent variations, without departing from the scope of the present teachings, using the methods and techniques disclosed above. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention are still within the scope of the protection of the technical solution of the present invention, unless the contents of the technical solution of the present invention are departed.

Claims (3)

1. A migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth is characterized by comprising the following steps:
(1) acquiring observation information LAT, LON and D of a certain migrant bird in a year period on a citizen data platform eBird, wherein the acquired observation information is stored in an array form and is expressed as follows:
Figure FDA0003142180480000011
wherein, LATjThe latitude, LON of the jth observation information of the migratory birdjLongitude of j-th observation of the waiting bird, DjN is the number of acquired observation information of the migratory bird on the j dayAmount, m is the number of days of the acquired observation information of the migratory bird, and m belongs to [1,366 ]],n≥m；
(2) If the repeated observation information data exists, deleting the same observation longitude and latitude information on the same date to reduce data set redundancy;
(3) converting longitude and latitude coordinates LAT and LON of the observed geographic position of the migratory bird into mercator coordinates which are marked as M-LAT and M-LON;
(4) to ensure the time continuity of the data, the observation date D missing in the original data is usedsAnd its corresponding observed geographic information LATs,LONsSelecting k nearest neighbor samples, performing linear interpolation by taking Euclidean distance as a distance function, recording the longitude and latitude and observation date of observation information obtained after interpolation as LM-LAT, LM-LON and LD, recording the number of bird observation information as N, and recording the number of days of bird observation information as M;
(5) in order to eliminate abnormal data points existing in the citizen observation data set, aiming at the characteristics of the space data set, an abnormal point detection algorithm based on space data point influence factors is adopted:
define LM-LAT, LM-LON as data set X ═ X1,X2,...XN}, set of weights
Figure FDA0003142180480000012
The spatial k distance neighborhood N (k, O) of the spatial object O can be represented as:
Figure FDA0003142180480000013
where k is the distance of a given spatial object o and dist (p, o, ω) is the distance of spatial objects p and o;
spatial object XiAnd XjThe distance of (d) can be defined as:
Figure FDA0003142180480000021
wherein，xik,xjkAre each Xi,XjF is a normalization function;
the average μ of the object to all neighbors in its neighborhood can be written as:
Figure FDA0003142180480000022
wherein N isK(p) space k distance neighborhood, N, representing object pK(p) | is the number of all neighbors in the neighborhood;
the spatial offset rate of the object p
Figure FDA0003142180480000023
Comprises the following steps:
Figure FDA0003142180480000024
spatial offset impact Rate of object p
Figure FDA0003142180480000025
Comprises the following steps:
Figure FDA0003142180480000026
the spatial offset impact factor of object p can be derived
Figure FDA0003142180480000027
Comprises the following steps:
Figure FDA0003142180480000028
and removing abnormal points with larger deviation influence factors in the observed data points to improve the data quality, and obtaining:
Figure FDA0003142180480000029
wherein r is the number of the observation information of the migratory birds obtained after the abnormal points are removed, and DLM-LAT and DLM-LON are the longitude and latitude of the observation information obtained by processing through an abnormal point detection algorithm;
(6) aggregating the observation data of the continuous seven days into the observation data of the first day of the seven days by adopting a sliding window idea to increase the stability of clustering, and recording the longitude and latitude of the observation information obtained by aggregating the sliding window idea as HDLM-LAT and HDLM-LON;
(7) clustering and clustering daily observing data of the migratory birds by adopting a Mean-Shift algorithm based on self-adaptive bandwidth, and generating the size of a Mean-Shift window by utilizing a bandwidth estimation algorithm, wherein the size is expressed by randomly selecting P samples from the observing data of the migratory birds to be classified, and for any sample point PiDefining the distance of each pair of samples in the unsupervised n-neighbor search result as:
M＝(m1,m2,...,mn)
selecting the maximum distance in M
Figure FDA0003142180480000031
Calculating the sum of the farthest pairing distances of the unsupervised n-neighbor search of the P samples, and recording as:
Figure FDA0003142180480000032
then selecting
Figure FDA0003142180480000033
As the final bandwidth, obtaining an unsupervised observation clustering result every day by using a Mean-Shift algorithm;
(8) according to the unsupervised observation clustering condition of each day, performing distance traversal on the clustering centers of two adjacent days by using a traversal algorithm based on minimum cost:
calculating the distance D between any two clustering centers in two adjacent days by using a tree structure hierarchical traversal method:
Figure FDA0003142180480000034
wherein (x)1,x2),(y1,y2) D is a coordinate under the mercator coordinate system of the clustering center and is in direct proportion to the cost;
obtaining the grouping condition of the population clustering center in the annual cycle;
Figure FDA0003142180480000035
wherein T represents the total group number of the migratory bird groups, which is equal to the number of the finally fitted tracks;
(9) carrying out nonlinear fitting on longitude and latitude of the observation grouped data of the migratory birds in an annual period and time respectively by utilizing a Generalized Additive Model (GAM), thereby obtaining the functional relation between the longitude and the latitude and the time respectively;
(10) and finding out the corresponding longitude and latitude coordinates on the same date, and connecting the geographic positions of all the dates to obtain the final migratory bird migration track reconstruction result.
2. The method according to claim 1, wherein the step (2) of deleting the same observed longitude and latitude information on the same date is specifically represented as: when any observation date D is selectedjWhen the temperature of the water is higher than the set temperature,
LATp≠LATqand LONp≠LONq
Wherein, LATp,LATqLatitude, LON of any two observation information of the j-th birdp,LONqThe longitude of any two pieces of observation information of the j-th bird.
3. The method of claim 1, wherein in step (4), k is 2, and the linear interpolation is formulated as follows:
Figure FDA0003142180480000041
Figure FDA0003142180480000042
Figure FDA0003142180480000043
wherein D ist,DlIs DsObservation date of k-neighbors, LATt,LATlIs LATsLatitude information of observation site of k-nearest neighbor, LONt,LONlIs LONsk-observation place longitude information of adjacent neighbors, wherein l is less than s and less than t;
merging the obtained interpolation into the data set without interpolation to obtain:
Figure FDA0003142180480000044
where M-365 or M-366.
CN202110737780.8A 2021-06-30 2021-06-30 Adaptive bandwidth-based bird waiting migration track reconstruction method Active CN113486940B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110737780.8A CN113486940B (en) 2021-06-30 2021-06-30 Adaptive bandwidth-based bird waiting migration track reconstruction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110737780.8A CN113486940B (en) 2021-06-30 2021-06-30 Adaptive bandwidth-based bird waiting migration track reconstruction method

Publications (2)

Publication Number Publication Date
CN113486940A true CN113486940A (en) 2021-10-08
CN113486940B CN113486940B (en) 2023-10-31

Family

ID=77936845

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110737780.8A Active CN113486940B (en) 2021-06-30 2021-06-30 Adaptive bandwidth-based bird waiting migration track reconstruction method

Country Status (1)

Country Link
CN (1) CN113486940B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014180241A (en) * 2013-03-19 2014-09-29 Tasada Kosakusho:Kk Migratory bird observation system and transmitter for the same
CN106940718A (en) * 2017-03-14 2017-07-11 山东师范大学 A kind of method, device and server for obtaining migratory bird moving path description
CN108268053A (en) * 2018-01-11 2018-07-10 北京航空航天大学 The autonomous cluster formation rotation control method of unmanned plane of imitative migratory bird evolution Snowdrift game
CN112396626A (en) * 2020-12-24 2021-02-23 西安费斯达自动化工程有限公司 Sequence image population identification method for monitoring migratory bird migration

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014180241A (en) * 2013-03-19 2014-09-29 Tasada Kosakusho:Kk Migratory bird observation system and transmitter for the same
CN106940718A (en) * 2017-03-14 2017-07-11 山东师范大学 A kind of method, device and server for obtaining migratory bird moving path description
CN108268053A (en) * 2018-01-11 2018-07-10 北京航空航天大学 The autonomous cluster formation rotation control method of unmanned plane of imitative migratory bird evolution Snowdrift game
US20190033893A1 (en) * 2018-01-11 2019-01-31 Beihang University UAV autonomous swarm formation rotation control method based on simulated migratory bird evolutionary snowdrift game
CN112396626A (en) * 2020-12-24 2021-02-23 西安费斯达自动化工程有限公司 Sequence image population identification method for monitoring migratory bird migration

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
乔慧捷;汪晓意;王伟;罗振华;唐科;黄燕;杨胜男;曹伟伟;赵新全;江建平;胡军华;: "从自然保护区到国家公园体制试点:三江源国家公园环境覆盖的变化及其对两栖爬行类保护的启示", 生物多样性, no. 02 *
胡中常;袁志湘;刘继群;覃晓春;廖汉军: "湖南屏风界候鸟种类与迁徙规律调查", 野生动物, vol. 28, no. 1 *

Also Published As

Publication number Publication date
CN113486940B (en) 2023-10-31

Similar Documents

Publication Publication Date Title
Aswin et al. Deep learning models for the prediction of rainfall
Kumar et al. Parallel k-means clustering for quantitative ecoregion delineation using large data sets
Sun et al. An overview of moving object trajectory compression algorithms
CN107688906B (en) Multi-method fused transmission line meteorological element downscaling analysis system and method
Lee et al. Precision, accuracy, and costs of survey methods for giraffe Giraffa camelopardalis
CN103235982A (en) BNM-based (Bayesian network model-based) fishery forecasting method
CN111242352A (en) Parking aggregation effect prediction method based on vehicle track
Zhao et al. Mastgn: Multi-attention spatio-temporal graph networks for air pollution prediction
CN114442623A (en) Agricultural machinery operation track field segmentation method based on space-time diagram neural network
CN113486940A (en) Migratory bird migration trajectory reconstruction method based on self-adaptive bandwidth
Fauchald et al. Arctic-breeding seabirds’ hotspots in space and time-A methodological framework for year-round modelling of environmental niche and abundance using light-logger data
Sambo et al. Integration of GPS and satellite images for detection and classification of fleet hotspots
Magnusson Uncertainty and the design of in-situ biodiversity-monitoring programs
Christina et al. Analysis of hard clustering algorithms applicable to regionalization
Moudgil et al. Groundwater levels estimation from GRACE/GRACE-FO and hydro-meteorological data using deep learning in Ganga River basin, India
Badr et al. Toward large-scale crop production forecasts for global food security
Yusof et al. Mapping frequent spatio-temporal wind profile patterns using multi-dimensional sequential pattern mining
Zhang et al. Dac-ml: domain adaptable continuous meta-learning for urban dynamics prediction
Feng et al. A novel method for multi-trajectory reconstruction based on LoMcT for avian migration in population level
Lucas et al. Exploring data quantity requirements for domain adaptation in the classification of satellite image time series
CN109508749B (en) Clustering analysis system and method based on depth knowledge expression
CN111125274A (en) Basic geographic information visualization method based on distributed computation
Tallón-Ballesteros Responses of climate indicators to droughts in SF bay area
Shirkey et al. Vocalization behavior of resident and migrant Virginia Rails (Rallus limicola) and Soras (Porzana carolina) in Northwestern Ohio, USA
Anand et al. Contextual anomaly detection in spatio-temporal data using locally dense regions

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant