CN108195767B - Estuary wetland foreign species monitoring method - Google Patents
Estuary wetland foreign species monitoring method Download PDFInfo
- Publication number
- CN108195767B CN108195767B CN201711416470.6A CN201711416470A CN108195767B CN 108195767 B CN108195767 B CN 108195767B CN 201711416470 A CN201711416470 A CN 201711416470A CN 108195767 B CN108195767 B CN 108195767B
- Authority
- CN
- China
- Prior art keywords
- data
- aerial vehicle
- unmanned aerial
- information
- satellite
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1793—Remote sensing
- G01N2021/1797—Remote sensing in landscape, e.g. crops
Abstract
The invention provides a estuary wetland exotic species monitoring method, which comprises the following steps of S1: selecting an unmanned aerial vehicle platform and a satellite remote sensing film source according to monitoring requirements; s2: setting a route of the unmanned aerial vehicle in the target monitoring area and setting corresponding flight parameters for the unmanned aerial vehicle; s3: setting a data acquisition task of the unmanned aerial vehicle; s4: executing a data acquisition task of the unmanned aerial vehicle to obtain data of the unmanned aerial vehicle; s5: matching geographic information of unmanned aerial vehicle data with satellite data to obtain a preprocessed satellite picture; s6: extracting plaque characteristic information in the preprocessed satellite picture; s7, screening according to the plaque characteristic information to obtain an optimal separability characteristic index; s8: and classifying the satellite data according to the optimal separability characteristic index and evaluating the precision of the result. The invention provides a estuary wetland exotic species monitoring method, which classifies invasive species vegetation by using remote sensing software and geographic information system software and has the advantages of high timeliness, high resolution, high accuracy, strong objectivity and low cost.
Description
Technical Field
The invention relates to the field of ecological environment assessment, in particular to a estuary wetland exotic species monitoring method.
Background
The method is a main tool for mastering the ecological environmental characteristics of the wetland, and can provide scientific basis and management suggestion for protecting wetland ecosystem organisms, evaluating benthic environmental quality and restoring the ecology of damaged habitats. Along with the continuous deepening of the globalization process, the biodiversity in the coastal/estuary wetland ecosystem is rapidly lost due to the biological invasion, and attention is urgently needed for monitoring and managing the distribution condition of non-native plants. However, the spatial distribution of invasive plants on a landscape scale has been far from being studied by means of traditional surveying alone. Fortunately, the numerous advantages of remote sensing technology provide new solutions to this problem.
The remote sensing technology has the advantages of wide coverage, various space and time scales, rich spectral information, flexible observation, convenient data acquisition and the like, has become the most cost-effective and efficient data acquisition means in wetland ecological environment monitoring, and plays a significant role in wetland planning, management and protection. Because the intertidal zone is influenced by human activities such as biological invasion, tidal action, grazing and the like, plant or ground object patches with small areas are easily generated, and the remote sensing image data source with high spatial resolution is required for accurately monitoring the spatial distribution and dynamic change of the plants or ground object patches. High spatial resolution data (VHR) are increasingly used to study and monitor the spatial distribution of invasive plant species, including Pleiades, QuickBird, IKONOS, IRS-P5, aerospace platforms, and almost all aerial images, with some spatial resolution even below l meters. On the other hand, due to the high similarity of spectral features of different plant species in the intertidal zone, it is necessary to acquire data at certain phenological stages and with appropriate spatial and spectral resolution, so that the separability is increased by the phenological information. It has been shown that the timing of data acquisition plays an important role, as plants are usually more different from the surrounding vegetation at some times of the vegetation season. Using a change monitoring approach, life cycle differences of certain species can be identified as compared to background vegetation. The timing of the acquisition of high spatial resolution remote sensing data is important since plant species may be better identified at certain climatic stages. However, acquisition of VHR satellite data is costly, and is generally constrained by the normal trajectory of the cloud and satellites. The satellite has low flexibility for moving in the air, needs planning in advance, is very expensive, and cannot be guaranteed not to be shielded by cloud layers through uncertain weather forecast. Therefore, the current selection of VHR data phenolics is limited.
In recent years, the unmanned aerial vehicle technology in China is becoming mature day by day, and the unmanned aerial vehicle has the advantages of portability, low cost, low loss, reusability, small risk, easiness in deployment and the like, and provides high flexibility for ground feature monitoring. In addition, the unmanned aerial vehicle can acquire the spatial characteristics of the ground features or the texture characteristics between the ground features and the adjacent ground features or the position information on a medium scale, the information breaks through the space-time limitation of field operation, more comprehensive information is provided for fine classification of the ground features, and then a classification algorithm is optimized. However, in the aspect of plant monitoring, classification is performed by means of images acquired by the unmanned aerial vehicle, and the defects that the spectral resolution of the images of the unmanned aerial vehicle is low and complicated processing is required due to geometric distortion and radiation irregularity are faced.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides the estuary wetland exotic species monitoring method, which classifies the vegetation of the invasive species by utilizing the existing mature remote sensing software and geographic information system software and has the advantages of high timeliness, high resolution, high information quantity, high field investigation efficiency, high accuracy, strong objectivity and low cost.
In order to achieve the above object, the present invention provides a estuary wetland exotecia monitoring method, comprising the steps of:
s1: selecting an unmanned platform and a satellite remote sensing film source according to monitoring requirements;
s2: setting at least one unmanned plane in a target monitoring area and setting corresponding flight parameters for the unmanned plane;
s3: setting a non-human-machine data acquisition task according to the climate data of the target monitoring area;
s4: executing the data acquisition task of the unmanned aerial vehicle to obtain data of the unmanned aerial vehicle;
s5: matching the geographic information of the unmanned aerial vehicle data with satellite data obtained from the satellite remote sensing film source to obtain a preprocessed satellite picture;
s6: extracting plaque characteristic information in the preprocessed satellite picture;
s7: screening according to the plaque characteristic information to obtain an optimal separability characteristic index;
s8: and classifying the satellite data according to the optimal separability characteristic index and evaluating the precision of the result.
Preferably, in the step S1, the satellite remote sensing film sources include Pleiades remote sensing high-resolution image source, QuickBird remote sensing high-resolution image source, and IKONOS remote sensing high-resolution image source.
Preferably, in the step S2, the flight parameters include a flight line, a shooting height, a shooting angle, a flying speed, and a shooting interval; setting the route, the shooting height and the shooting angle according to the satellite data of the satellite remote sensing film source; setting the flight speed according to the flight path; setting the shooting interval according to the flying speed, wherein adjacent pictures shot by the unmanned aerial vehicle according to the shooting interval have an image overlapping area; and obtaining the cross-sectional distance d of the intertidal zone of the target monitoring area.
Preferably, the step of S3 further comprises the steps of:
s31: acquiring a tide table of the target monitoring area;
s32: acquiring the high tide level, the low tide level and the tide time of the target monitoring area in the high tide period according to the tide table;
s33: determining a shooting execution time period of the unmanned aerial vehicle according to the farthest flight distance D of the unmanned aerial vehicle and the cross section distance D of the intertidal zone to form a data acquisition task of the unmanned aerial vehicle;
when D is larger than D, the shooting execution time period of the unmanned aerial vehicle is positioned before the tide time of the high tide level, and the time length of the shooting execution time period of the unmanned aerial vehicle is larger than or equal to the round trip time of the unmanned aerial vehicle shooting along the route;
when D is less than or equal to D, a worker needs to carry the unmanned aerial vehicle to move to a position away from the water and the distance D' from the water is less than the farthest flight distance D, then the unmanned aerial vehicle enters the shooting execution time period, and the duration of the shooting execution time period of the unmanned aerial vehicle is greater than or equal to the sum of the round trip time of the unmanned aerial vehicle shooting along the air route and the round trip time of the worker.
Preferably, the step of S5 further comprises the steps of:
s51: preprocessing the satellite data by band fusion and geometric correction to obtain preprocessed remote sensing data;
s52: importing the unmanned aerial vehicle data through geographic information system software;
s53: vectorizing the unmanned aerial vehicle data, wherein the vectorized unmanned aerial vehicle data comprises photo data, first GPS information and plant species information, and each first vector point of the photo data is associated with the first GPS information and the plant species information;
s54: the preprocessed remote sensing data comprise satellite image data, second GPS information, unknown plant species spectrum information and spatial information; and matching each second vector point of the satellite image data with each associated information of the first vector point of the photo data by utilizing the consistency of the first GPS information and the second GPS information, so that the unknown plant species spectrum information is matched with the plant species information.
Preferably, in the step S6, the range of the same plant in the preprocessed satellite picture is obtained by using the segmentation function of the ecographing developer software, and a plurality of patches are formed in the preprocessed satellite picture according to the plant species; assigning and associating the corresponding plant species information for the plaque by utilizing a classification function of eCognitionDeveloper software; forming plaque feature information, the plaque feature information comprising: plaque spectral information, plaque shape information, and plaque location information.
Preferably, the step of S7 further comprises the steps of:
outputting statistical data of the plaque characteristic information by using an output function in the eCooginionDeveloper software;
and analyzing the statistical data through multivariate variance, screening out the optimal separability characteristic index and determining the threshold range of each type of vegetation.
Preferably, the step of S8 further comprises the steps of:
establishing a classification algorithm rule set through the eCoginationDeveloper software based on the optimal separability characteristic index;
classifying the satellite data by using a supervision classification method of an algorithm set in the eCognitionDeplastics software to obtain a classification pixel set;
and comparing the classified pixel set with a reference pixel set, and evaluating the precision of the satellite data.
Due to the adoption of the technical scheme, the invention has the following beneficial effects:
firstly, the invention provides an accurate monitoring method suitable for vegetation species by utilizing the space-time convenience of unmanned aerial vehicle data and the function of accurately positioning vegetation species information, combining the abundant spectrum and space texture information of high-spatial resolution remote sensing data and the handlability of available mature classification software, and has advantages in monitoring selected invasive species. The method overcomes the difficult problems that the monitoring is only carried out through high-resolution remote sensing data in a wetland area, the monitoring is limited by weather, tide and cost, the available data is less, and the invasive vegetation is difficult to accurately monitor through enough phenological or spectral information. An optimal combination and balance between spatial/spectral/temporal resolution, cost, accuracy is achieved.
Secondly, the implementation of unmanned aerial vehicle sampling still overcomes the difficulty of sampling at wetland spot in the field, and unmanned aerial vehicle is easy to operate, and coverage is big, reduces artifical field work's time and work load, improves the efficiency of investigation in the field, not only can obtain the large amount of sample volume that artificial sampling can not reach, and can also obtain the plaque information on the mesoscale that it can not obtain, improves classification accuracy.
Drawings
Fig. 1 is a flow chart of a estuary wetland exotic species monitoring method according to an embodiment of the invention.
Detailed Description
The following description of the preferred embodiment of the present invention, in accordance with the accompanying drawings of which 1 is presented to enable a better understanding of the invention as to its functions and features.
Referring to fig. 1, a method for monitoring alien species in estuary wetland according to an embodiment of the present invention includes the steps of:
s1: and selecting a unmanned platform and a satellite remote sensing film source according to the monitoring requirement.
In the embodiment, a Phantom 3 unmanned aerial vehicle of the Dajiang Corporation (DJI Corporation) is selected, which is a small-sized, battery-powered, high-definition camera-mounted and relatively cheap unmanned aerial vehicle, and can automatically set route planning, shooting angle, height, shooting interval and flying speed; the acquired photos integrate built-in GPS geographic information and can be matched with satellite data in a non-difference mode.
The satellite remote sensing film source comprises a Pleiades remote sensing high-resolution image source, a Quickbird remote sensing high-resolution image source, an IKONOS remote sensing high-resolution image source and the like; the specific type of the remote sensing high-resolution image source can also be not limited by the existing other remote sensing high-resolution image sources.
In addition, when selecting the sheet source, the proper phenological period sheet source is selected under the selectable sheet source as much as possible, and the phenological period time with larger difference between the monitored vegetation species and the surrounding vegetation, such as the budding period, the flowering period or the withering period, is selected as much as possible.
S2: at least one unmanned plane is set in a target monitoring area, and corresponding flight parameters are set for the unmanned plane.
The flight parameters comprise a flight line, shooting height, shooting angle, flight speed and shooting interval; setting a route, a shooting height and a shooting angle according to satellite data of a satellite remote sensing film source; setting the flight speed according to the flight path; and setting a shooting interval according to the flying speed, wherein adjacent pictures shot by the unmanned plane according to the shooting interval have an image overlapping area. The intertidal zone cross-sectional distance d of the target monitoring area can be obtained when setting the route.
In the embodiment, the route can be set according to the heterogeneity characteristic of the remote sensing image landscape in the satellite data of the satellite remote sensing film source; setting a shooting height according to the recognizable degree of the species and the size of the plaque in the picture; setting a shooting angle according to the vegetation canopy characteristics; setting the flying speed according to the flying distance of each flight path; and setting the shooting interval according to the flying speed, so that the overlapping and diagonal lines of the photos are respectively between 80% and 85% of the height and width of the images, and ensuring that the photos are sufficiently overlapped to realize uniform coverage of a sampling area.
S3: setting a task of unmanned data acquisition according to the climate data of the target monitoring area, further comprising the following steps:
s31: acquiring a tide table of a target monitoring area;
s32: acquiring the high tide level, the low tide level and the tide time of a target monitoring area in a high tide period according to a tide table, and acquiring the cross section distance d of an intertidal zone of the target monitoring area according to the high tide level, the low tide level and the tide time;
s33: determining a shooting execution time period of the unmanned aerial vehicle according to the farthest flight distance D of the unmanned aerial vehicle and the cross section distance D of the intertidal zone, and forming an unmanned aerial vehicle data acquisition task;
when D is larger than D, the shooting execution time period of the unmanned aerial vehicle is positioned before the tide time of the high tide level, and the duration of the shooting execution time period of the unmanned aerial vehicle is more than or equal to the round trip time of the unmanned aerial vehicle shooting along the air route;
and when the D is less than or equal to D, the worker needs to carry the unmanned aerial vehicle to move until the distance D' from the water to the water side is less than the farthest flight distance D and then enters the shooting execution time period of the unmanned aerial vehicle, and the duration of the shooting execution time period of the unmanned aerial vehicle is more than or equal to the sum of the round trip time of the unmanned aerial vehicle shooting along the air route and the round trip time of the worker.
S4: and executing the data acquisition task of the unmanned aerial vehicle to obtain the data of the unmanned aerial vehicle.
S5: matching geographic information of unmanned aerial vehicle data with satellite data obtained from a satellite remote sensing film source to obtain a preprocessed satellite picture; it further comprises the steps of:
s51: preprocessing satellite data by band fusion and geometric correction to obtain preprocessed remote sensing data;
s52: importing unmanned aerial vehicle data through geographic information system software; in the embodiment, the geographic information system software adopts ArcGIS software;
s53: vectorizing unmanned aerial vehicle Data through a Geobagged phones to point functional module of Data Management in ArcGIS software, wherein the vectorized unmanned aerial vehicle Data comprises photo Data, first GPS information and plant species information, and each first vector point of the photo Data is associated with the first GPS information and the plant species information.
S54: the preprocessed remote sensing data comprise satellite image data, second GPS information, unknown plant species spectrum information and spatial information; and matching each second vector point of the satellite image data with each information associated with the first vector point of the photo data by utilizing the consistency of the first GPS information and the second GPS information, so that the unknown plant species spectrum information is matched with the plant species information.
By the steps, the vegetation name, the spectral information and the spatial information of a certain point on the satellite image data can be obtained. And S7, extracting the known vegetation species sampling points in the satellite image data, the spectral information and the spatial information characteristics of the patches where the sampling points are located by using remote sensing classification software, and analyzing.
The method comprises the steps of preprocessing high-spatial-resolution remote sensing data, namely satellite data, and converting original data into usable data with high definition and accurate position by band fusion and geometric correction. The steps can be realized by a band fusion and geometric correction functional module of the existing remote sensing software.
Since the high spatial resolution remote sensing data comprises multispectral wave bands (for example, P star is 2m resolution) and panchromatic wave bands (for example, P star is 0.5m resolution). The multispectral wave band product provides richer ground object spectrum information than a panchromatic wave band product, can form a color synthetic image, is beneficial to the identification of the ground object, but has lower spatial resolution than the full-color wave band. The full-color waveband single waveband is a gray picture displayed on the picture, and the picture has high spatial resolution but cannot display the colors of the ground objects. And fusing the panchromatic waveband and the multispectral waveband images to obtain an image which has the high resolution of the panchromatic image and the color information of the multiband image.
Geometric correction generally refers to correcting and eliminating the image of a remote sensing image through a series of mathematical models, and due to the influence of factors such as the attitude, the height, the speed and the earth rotation of an aircraft, the image generates geometric distortion relative to a ground target, the distortion is expressed as extrusion, distortion, stretching, deviation and the like of the actual position of a pixel relative to the ground target, and the error correction performed on the geometric distortion is called geometric correction. Geometric fine correction: the geometric correction is carried out by using control points, which is to use a mathematical model to approximately describe the geometric distortion process of the remote sensing image, use some corresponding points (namely control point data pairs) between the distorted remote sensing image and a standard map to obtain the geometric distortion model, and then use the model to carry out the geometric distortion correction.
The unmanned aerial vehicle is internally provided with the GPS module, and the GPS positioning data can be obtained more quickly and accurately as long as more than 3 satellite signals can be received at the same time, and is recorded in EXIF information of the photo.
S6: and forming a plurality of plaques in the preprocessed satellite picture according to the plant types, and forming the plaque characteristic information of the plaques.
Obtaining the range of the same plant in the preprocessed satellite picture by utilizing the segmentation function of eCognitionDeveloper software, and forming a plurality of plaques in the preprocessed satellite picture according to the plant type; assigning and associating corresponding plant type information for the plaque by utilizing a classification function of eCognitionDeveloper software; forming plaque characteristic information, wherein the plaque characteristic information comprises: plaque spectral information, plaque shape information, and plaque location information.
S7: and screening according to the plaque characteristic information to obtain the optimal separability characteristic index.
Outputting statistical data of the plaque characteristic information by using an output (Export) function in eCooginion developer software; and screening out the optimal separability characteristic index and determining the threshold range of each type of vegetation through multivariate analysis of variance statistical data.
In this embodiment, in order to determine whether different feature information of different vegetation type patches has significance differences, the operation can be realized through the function module of various statistical software by analyzing statistical data through multivariate variance analysis.
In the statistical analysis, the mean, standard error and 95% confidence interval range for different characteristic parameters of different vegetation types are obtained, and these ranges constitute the threshold range of each type of vegetation under the characteristic parameters with significant differences.
S8: and classifying the satellite data according to the optimal separability characteristic index and evaluating the precision of the result.
Firstly, establishing a classification algorithm rule set through eCoginationDeveloper software based on the optimal separability characteristic index; in the embodiment, the classification algorithm rule set is established by using the eCogination Developer software, and only all the ground feature class names need to be established first, and then the supervision classification algorithm used under each class of ground features needs to be established; such as a minimum proximity classification algorithm; and then the determined optimization characteristic parameters are selected and added into each type of ground object category algorithm.
Secondly, classifying the satellite data by using a supervision classification method of an algorithm set in eCognitionDeplastics software to obtain a classification pixel set; the supervised classification is a technology for classifying images to be classified by establishing a statistical recognition function as a theoretical basis and according to a typical sample training method, namely, according to samples provided by a known training area, by selecting characteristic parameters, solving the characteristic parameters as a decision rule and establishing a discriminant function to classify the images to be classified, and is a method for pattern recognition. In supervision and classification, the selection of characteristic parameters is very important, so the purpose of the step is to analyze and determine the optimal characteristic parameters capable of distinguishing different vegetation types;
and finally, comparing the classified pixel set with a reference pixel set, and evaluating the precision of the satellite data.
The method adopted by the invention is applied to monitoring invasive plant spartina alterniflora in the wetland at the estuary of the Yangtze river as an example:
the research position is in the largest island of the Yangtze estuary, namely the east beach at the most east end of the Chongming island, the region is in the national natural protection area of the Chongming east beach, and the region is a typical coastal/estuary wetland. In order to meet the requirement of land, spartina alterniflora was introduced into China in 1979, is a salt-tolerant and flood-tolerant plant suitable for the growth of intertidal zones of the beach, and is mainly used for protecting the beach dykes, promoting silt and making land, improving soil and the like. Spartina alterniflora belongs to the genus Spartina of the family Gramineae, is a perennial tall herbaceous plant, is robust and tall, and has an average plant height of about 1.5 m. The spartina alterniflora has a series of mechanisms which are beneficial to the survival and the diffusion of the population of the spartina alterniflora, inhibits the growth of other plants, also causes the shellfish to have difficulty in moving in a dense spartina alterniflora beach, even suffocate and die, threatens the food sources of fishes and birds, reduces the biodiversity of beaches, seriously destroys the natural balance of the ecological sensitive area of the Chongming east beach, generates serious ecological and evolutionary consequences for the local area, changes the food net and the nutritional structure of the whole ecological system, causes many original fishes and migratory birds to lose nutritional support, and faces danger.
The Chongming east Tansta plant community has the characteristics of few dominant plant types, simple community types and the like, and is favorable for remote sensing monitoring research. Although there are about 96 higher plants in Chongmingtong beach, the dominant species are mainly Phragmites communis (Phragmitis), Spartina alterniflora (Spartinaalterniflora), Carexscabrifolia (Carexscabrifolia), Scirpus maritima (Scirpus marique). In geographical distribution, the vegetation has a certain banding phenomenon, namely, the vegetation gradually changes from a low-tide plain beach into a community taking scirpus maritima as a dominant species and having scirpus maritima and carex latifolius therebetween, and then the vegetation replaces the medium-tide and the high-tide level into a community mainly comprising reeds or spartina alterniflora and forms a situation of mutual competition. The growth height of bulrush and spartina alterniflora can reach 1.5-2.0 m, while the growth height of scripus triqueter, carex fuscus and scirpus maritima is about 0.3-0.7 m.
The first step is as follows: and selecting an unmanned aerial vehicle platform and a satellite remote sensing film source.
In this embodiment, a Phantom 3 unmanned aerial vehicle of the great jiang company (DJI Corporation) is selected, for a satellite remote sensing film source, a mainstream Pleiades star remote sensing high resolution image source is selected, the spatial resolution is 0.5m, according to field observation, superior species such as reed, sparrow, scirpus maritima and sedum glaucophyllum of chongdongtang exhibit certain phenological differences, particularly, the sequence of emergence period in spring and withering period in autumn is around, and according to the quality (angle and cloud amount) of the selectable images, the remote sensing image in 2015 10 months is finally selected as the target interpretation image. In the time period, the 10 middle ten days of reed is the full bloom period; the terminal 7 months of spartina alterniflora is the full bloom stage, and the middle 10 months is the seed setting stage; carex incarnata and scirpus maritima begin to wither in 10 middle ten days.
The second step is that: and setting a route of the unmanned aerial vehicle in the target monitoring area and setting corresponding flight parameters for the unmanned aerial vehicle.
During the flight mission of the unmanned aerial vehicle, air routes are set according to the heterogeneity characteristics of remote sensing image landscape, three flight stations are set from south to north, three air routes are set in three directions of each station, and 9 air routes are shared, so that all the air routes are distributed in the whole research area range more uniformly; setting a shooting height according to the recognizable degree (the size of the patch) of the species, wherein the diameter range of most of the patches of the plantlets is about 10-30 m according to field investigation experience, the flight height is equivalent to the diameter range of the patches according to actual flight test, and the recognizable degree of the shot pictures is 15-20 m; setting a shooting angle according to the height of a vegetation canopy and the type of vegetation, wherein dominant plants of the east beach wetland are herbaceous plants, the leaf canopy is thin and narrow, the plant height is mostly in the range of about 0.5-2 m, if the vegetation type is difficult to distinguish by orthographic overlooking shooting, the shooting angle is set to be 30-45 degrees in elevation angle, and the type of the plants in the photo can be well distinguished; the flight speed is set according to the distance of each flight path, the flight speed is set to be 10m/s according to the test because the distance of each flight path is about 1 km-2 km, the shooting interval is 2 s/piece, the photos can be overlapped, the diagonal lines are respectively 80% -85% of the height and the width of the image, the sufficient overlapping between the photos is ensured, and the uniform coverage of a sampling area is realized.
The third step: and setting a task of unmanned data acquisition according to the climate data of the target monitoring area.
Inquiring a tide table of the surveyed wetland area to obtain the high tide level, the low tide level and the tide time of the high tide period of the area; and determining the time for logging in and withdrawing the operating personnel and the executable time period for aerial photography of the low-altitude aircraft of the unmanned aerial vehicle according to the high tide level, the low tide level and the tide time of the high tide period and the altitude of the logging in and withdrawing areas.
The fourth step: and executing the data acquisition task of the unmanned aerial vehicle to obtain the data of the unmanned aerial vehicle.
The fifth step: and matching the geographic information of the unmanned aerial vehicle data with satellite data obtained from a satellite remote sensing film source to obtain a preprocessed satellite picture.
First, the multiband data and panchromatic band data of remote sensing image data (satellite data) are fused to make the resolution 0.5 m. The geometric correction of polynomial coefficients is carried out by using the topographic map of the country 1: 2000. Specifically, the method comprises the following steps: geometric correction is carried out on the purchased original high-resolution remote sensing data product, 5-10 pairs of control points (generally, road intersection points with high recognition degree, bridges or mark houses and the like) on the same position on the topographic map and the remote sensing image data are found, a formula coefficient with the lowest error is calculated by utilizing a polynomial coefficient calculation method through respective coordinates (XY values) of the control points, and all pixel points on the remote sensing image are converted through the formula to complete the geometric correction of the whole map.
Through geometric correction, the satellite data and the research area and the verification point in the unmanned aerial vehicle photo can be matched in the same area. Then, importing unmanned aerial vehicle photo GPS information through ArcGIS software to be matched with high spatial resolution remote sensing Data (VHR Data), converting the GPS information based on EXIF in the photo into a vector point format by using a Geotaggedphos to point tool of Data Management in the ArcGIS software, and tracing original photo vegetation information through an identity query tool, thereby adding vegetation species information to the vector point Data, wherein the species attribute information name is consistent with the category name of the classification process of the eCoginationDeplastics software.
And a sixth step: and extracting plaque characteristic information in the preprocessed satellite picture.
By utilizing the functions of Segmentation and Classification (Classification) in the eCoginationDeplastics software, the spectrum, the shape and the position information of the plaque of the invasive species and other species where the unmanned aerial vehicle sample point is located are extracted.
In the segmentation classification algorithm, multi-Scale segmentation (multi-resolution segmentation) is firstly carried out on the VHR remote sensing image (high spatial resolution remote sensing data), and the optimal Scale (Scale) which accords with the characteristics of most plant plaques is determined; under the optimal scale, a Chessboard segmentation method (chess board segmentation) is used to obtain the minimum segmentation object which is the same as the sample point and has the category information, namely, the sample point plaque used in the analysis in the following steps. Classifying the sampling point pattern spots through the attribute (Class _ name) of the sampling points by using a classification algorithm (assignment by the systematic layer); and continuing to Assign the sample patches overlapped with the sample point objects to the same categories as the sample points by using a classification algorithm (Assign Class) (by adding an attribute limiting condition: a border to Class attribute in Class-related features), and obtaining the uniform patches where the sample points of all known vegetation species are located.
The plaque spectrum, spatial features are selected by the plaque feature selection function in the ecognitiondeviloper software: adding an exponential characteristic instruction required for creating in the Feature View interface, such as: NDVI, layer means/stationary value, geometry-extension-area/length/width, position-distance to the scene coder, geometry-shape-dimension/binder index/compatibility/main-orientation/shape index, texture-layer value texture base sub-object, texture-shape value texture base sub-object, to obtain the relevant spatial feature value of the known sample point patch.
The seventh step: and screening according to the plaque characteristic information to obtain the optimal separability characteristic index.
Setting level, class filter and Features by using an Export object statistics algorithm (Export objects statistics) in eCoginationDeplastics software, outputting statistical data of each attribute information (spectrum, shape and position) of each plant patch, and then performing statistics, re-analysis and comparison to obtain effective Features for different classes.
And performing multivariate analysis of variance (analysis > general linear model > multivariate) through SPSS software, performing significance test on the difference of sample mean values of the plant plaque attributes of different types, screening out the optimal separability characteristic indexes, and determining the threshold range of the indexes for distinguishing each type of vegetation.
Eighth step: and evaluating the precision of the satellite data according to the optimal separability characteristic index.
Firstly, converting the image objects extracted in the fourth step, namely Classified sample plaques into Classified sub-samples (Classified image objects to samples); and establishing a classification algorithm rule set in eCognition Developer software based on the obtained optimal separability characteristic index, and performing supervision and classification on the whole VHR remote sensing image according to the algorithm set.
An error matrix (error matrix) method is a precision evaluation method, and is a precision evaluation method suitable for supervised classification by using sample characteristics. Is a comparison array used for representing the number of pixels classified into a certain category and the number of pixels classified into the category through ground inspection. Typically, the columns in the array represent reference data and the rows represent category data that is categorized from the telemetry data. The number of pixels and the percentage indicate two. From the error matrix overall classification accuracy, Kappa coefficient, misclassification error, missing classification error, drawing accuracy of each class and user accuracy, the accuracy and reliability of classification are determined by these accuracy values.
While the present invention has been described in detail and with reference to the embodiments thereof as illustrated in the accompanying drawings, it will be apparent to one skilled in the art that various changes and modifications can be made therein. Therefore, certain details of the embodiments are not to be interpreted as limiting, and the scope of the invention is to be determined by the appended claims.
Claims (3)
1. A estuary wetland exotic species monitoring method comprises the following steps:
s1: selecting an unmanned platform and a satellite remote sensing film source according to monitoring requirements;
when selecting a sheet source, selecting the sheet source for monitoring the phenological period time with larger difference between the vegetation species and the surrounding vegetation;
s2: setting at least one unmanned plane in a target monitoring area and setting corresponding flight parameters for the unmanned plane; setting the route according to the heterogeneity characteristic of the remote sensing image landscape in the satellite data of the satellite remote sensing film source;
s3: setting a non-human-machine data acquisition task according to the climate data of the target monitoring area;
s4: executing the data acquisition task of the unmanned aerial vehicle to obtain data of the unmanned aerial vehicle;
s5: matching the geographic information of the unmanned aerial vehicle data with satellite data obtained from the satellite remote sensing film source to obtain a preprocessed satellite picture;
s6: forming a plurality of plaques in the preprocessed satellite picture according to plant species, and forming plaque characteristic information of the plaques;
s7: screening according to the plaque characteristic information to obtain an optimal separability characteristic index;
s8: classifying the satellite data according to the optimal separability characteristic index and evaluating the precision of the result;
in the step of S1, the satellite remote sensing film source comprises a Pleiades remote sensing high-resolution image source, a Quickbird remote sensing high-resolution image source and an IKONOS remote sensing high-resolution image source;
in the step S2, the flight parameters include a flight line, a shooting height, a shooting angle, a flight speed, and a shooting interval; setting the route, the shooting height and the shooting angle according to the satellite data of the satellite remote sensing film source; setting the flight speed according to the flight path; setting the shooting interval according to the flying speed, wherein adjacent pictures shot by the unmanned aerial vehicle according to the shooting interval have an image overlapping area; obtaining the cross-sectional distance d of the intertidal zone of the target monitoring area;
the step of S7 further includes the steps of:
outputting statistical data of the plaque characteristic information by using an output function in eCooginionDeveloper software;
analyzing the statistical data through multivariate variance, screening out the optimal separability characteristic index and determining the threshold range of each type of vegetation;
the step of S5 further includes the steps of:
s51: preprocessing the satellite data by band fusion and geometric correction to obtain preprocessed remote sensing data;
s52: importing the unmanned aerial vehicle data through geographic information system software;
s53: vectorizing the unmanned aerial vehicle data, wherein the vectorized unmanned aerial vehicle data comprises photo data, first GPS information and plant species information, and each first vector point of the photo data is associated with the first GPS information and the plant species information;
s54: the preprocessed remote sensing data comprise satellite image data, second GPS information, unknown plant species spectrum information and spatial information; matching each second vector point of the satellite image data with each associated information of the first vector point of the photo data by utilizing the consistency of the first GPS information and the second GPS information, so that the unknown plant species spectrum information is matched with the plant species information;
the step of S8 further includes the steps of:
establishing a classification algorithm rule set through the eCoginationDeveloper software based on the optimal separability characteristic index;
classifying the satellite data by using a supervision classification method of an algorithm set in the eCognitionDeplastics software to obtain a classification pixel set;
and comparing the classified pixel set with a reference pixel set, and evaluating the precision of the satellite data.
2. The estuary wetland alien species monitoring method according to claim 1, wherein the step S3 further comprises the steps of:
s31: acquiring a tide table of the target monitoring area;
s32: acquiring the high tide level, the low tide level and the tide time of the target monitoring area in the high tide period according to the tide table;
s33: determining a shooting execution time period of the unmanned aerial vehicle according to the farthest flight distance D of the unmanned aerial vehicle and the cross section distance D of the intertidal zone to form a data acquisition task of the unmanned aerial vehicle;
when D is larger than D, the shooting execution time period of the unmanned aerial vehicle is positioned before the tide time of the high tide level, and the time length of the shooting execution time period of the unmanned aerial vehicle is larger than or equal to the round trip time of the unmanned aerial vehicle shooting along the route;
when D is less than or equal to D, a worker needs to carry the unmanned aerial vehicle to move to a position away from the water and the distance D' from the water is less than the farthest flight distance D, then the unmanned aerial vehicle enters the shooting execution time period, and the duration of the shooting execution time period of the unmanned aerial vehicle is greater than or equal to the sum of the round trip time of the unmanned aerial vehicle shooting along the air route and the round trip time of the worker.
3. The estuary wetland exotspecies monitoring method according to claim 2, wherein in the step S6, the range of the same plant in the pre-processing satellite picture is obtained by using the segmentation function of the ecognition developer software, and a plurality of plaques are formed in the pre-processing satellite picture according to the plant species; assigning and associating the corresponding plant species information for the plaque by utilizing the classification function of the eCognitionDeveloper software; forming plaque feature information, the plaque feature information comprising: plaque spectral information, plaque shape information, and plaque location information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711416470.6A CN108195767B (en) | 2017-12-25 | 2017-12-25 | Estuary wetland foreign species monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711416470.6A CN108195767B (en) | 2017-12-25 | 2017-12-25 | Estuary wetland foreign species monitoring method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108195767A CN108195767A (en) | 2018-06-22 |
| CN108195767B true CN108195767B (en) | 2020-07-31 |
Family
ID=62583794
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711416470.6A Active CN108195767B (en) | 2017-12-25 | 2017-12-25 | Estuary wetland foreign species monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108195767B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110390255A (en) * | 2019-05-29 | 2019-10-29 | 中国铁路设计集团有限公司 | High-speed rail environmental change monitoring method based on various dimensions feature extraction |
| CN111047566B (en) * | 2019-12-04 | 2023-07-14 | 昆明市滇池高原湖泊研究院 | Method for carrying out aquatic vegetation annual change statistics by unmanned aerial vehicle and multispectral satellite image |
| CN111553244B (en) * | 2020-04-24 | 2023-05-26 | 中国电建集团成都勘测设计研究院有限公司 | Water and soil conservation monitoring method based on automatic positioning and orientation technology |
| CN111767643A (en) * | 2020-06-03 | 2020-10-13 | 中国水产科学研究院东海水产研究所 | Method for determining distribution range of biological subject environmental factors |
| CN112767628A (en) * | 2020-12-18 | 2021-05-07 | 交通运输部水运科学研究所 | Port foreign organism intrusion disaster tracking and monitoring system |
| CN114414701B (en) * | 2022-01-28 | 2023-06-13 | 辽宁中医药大学 | Preparation method of myrobalan acid in myrobalan meat and content determination method thereof |
| CN115810155B (en) * | 2023-01-18 | 2023-04-18 | 中关村睿宸卫星创新应用研究院 | Tidal wetland classification method |
| CN115834845A (en) * | 2023-02-24 | 2023-03-21 | 山东慧创信息科技有限公司 | Unmanned aerial vehicle land data identification system based on thing networking |
| CN116625354B (en) * | 2023-07-21 | 2023-09-29 | 山东省国土测绘院 | High-precision topographic map generation method and system based on multi-source mapping data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104915674A (en) * | 2014-10-24 | 2015-09-16 | 北京师范大学 | Landsat8 and MODIS fusion-construction high space-time resolution data identification autumn grain crop method |
| CN106915462A (en) * | 2017-02-14 | 2017-07-04 | 福建兴宇信息科技有限公司 | Forestry pests ＆ diseases intelligent identifying system based on multi-source image information |
| CN107389036A (en) * | 2017-08-02 | 2017-11-24 | 珠江水利委员会珠江水利科学研究院 | A kind of large spatial scale vegetation coverage computational methods of combination unmanned plane image |
| CN107478611A (en) * | 2017-08-02 | 2017-12-15 | 珠江水利委员会珠江水利科学研究院 | A kind of method for calculating the exposed rate of rock |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103528570A (en) * | 2013-10-23 | 2014-01-22 | 环境保护部卫星环境应用中心 | Spatial distribution acquiring method and system for spartina alterniflora in northern sea area of Guangxi |
| CN104463470B (en) * | 2014-12-10 | 2017-06-20 | 中国水产科学研究院东海水产研究所 | Intertidal Habitats investigation method based on unmanned plane |
-
2017
- 2017-12-25 CN CN201711416470.6A patent/CN108195767B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104915674A (en) * | 2014-10-24 | 2015-09-16 | 北京师范大学 | Landsat8 and MODIS fusion-construction high space-time resolution data identification autumn grain crop method |
| CN106915462A (en) * | 2017-02-14 | 2017-07-04 | 福建兴宇信息科技有限公司 | Forestry pests ＆ diseases intelligent identifying system based on multi-source image information |
| CN107389036A (en) * | 2017-08-02 | 2017-11-24 | 珠江水利委员会珠江水利科学研究院 | A kind of large spatial scale vegetation coverage computational methods of combination unmanned plane image |
| CN107478611A (en) * | 2017-08-02 | 2017-12-15 | 珠江水利委员会珠江水利科学研究院 | A kind of method for calculating the exposed rate of rock |
Non-Patent Citations (3)
| Title |
|---|
| Estimating Spartina alterniflora fractional vegetation cover and aboveground biomass in a coastal wetland using SPOT6 satellite and UAV data;Zhou Zaiming等;《Aquatic Botany》;20171024;第144卷;第38-45页 * |
| Species classification using Unmanned Aerial Vehicle (UAV)-acquired high spatial resolution imagery in a heterogeneous grassland;Lu Bing等;《ISPRS Journal of Photogrammetry and Remote Sensing》;20170329;第128卷;第73-85页 * |
| 基于无人机影像的农情遥感监测应用;王利民等;《农业工程学报》;20130930;第29卷(第18期);第136-145页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108195767A (en) | 2018-06-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108195767B (en) | Estuary wetland foreign species monitoring method | |
| Mucher et al. | Land cover characterization and change detection for environmental monitoring of pan-Europe | |
| Al-Ali et al. | A comparative study of remote sensing classification methods for monitoring and assessing desert vegetation using a UAV-based multispectral sensor | |
| CN109977991A (en) | Forest resourceies acquisition method based on high definition satellite remote sensing | |
| Zhu et al. | Estimating and mapping mangrove biomass dynamic change using WorldView-2 images and digital surface models | |
| Gómez‐Sapiens et al. | Improving the efficiency and accuracy of evaluating aridland riparian habitat restoration using unmanned aerial vehicles | |
| CN114581784A (en) | Construction method of long-time-sequence yearly mangrove remote sensing monitoring product | |
| Zhang et al. | UAV‐derived imagery for vegetation structure estimation in rangelands: validation and application | |
| Udin et al. | Land use and land cover detection by different classification systems using remotely sensed data of Kuala Tiga, Tanah Merah Kelantan, Malaysia | |
| Coulter et al. | Comparison of high spatial resolution imagery for efficient generation of GIS vegetation layers | |
| CN115760885B (en) | High-closure-degree wetland forest parameter extraction method based on consumer-level unmanned aerial vehicle image | |
| Cameron et al. | Temporal measurement of the loss of native pinewood in Scotland through the analysis of orthorectified aerial photographs | |
| Macedo et al. | Land use and land cover mapping in detailed scale: a case study in Santa Rosa de Lima-SC | |
| Whiteside et al. | Vegetation map for Magela Creek floodplain using WorldView-2 multispectral image data | |
| Su et al. | Evaluation of eelgrass beds mapping using a high-resolution airborne multispectral scanner | |
| Wilkinson et al. | Change detection techniques for use in a statewide forest inventory program | |
| DiGiacomo et al. | Considerations and tradeoffs of UAS-based coastal wetland monitoring in the Southeastern United States | |
| GASTELLU-ETCHEGORRY et al. | Computer-assisted land cover mapping with SPOT in Indonesia | |
| De Giglio et al. | Satellite and unmanned aerial vehicle data for the classification of sand dune vegetation | |
| Martin | Determining estuarine seagrass density measures from low altitude multispectral imagery flown by remotely piloted aircraft | |
| Pepe | Use of digital aerial photogrammetry sensors for land cover classification | |
| Kumar et al. | Accuracy assessment of land cover classification and change detection using remote sensing and GIS | |
| Conrad et al. | Mapping crop distribution in administrative districts of southwest Germany using multi-sensor remote sensing data | |
| Rudra et al. | Spatiotemporal changing pattern of land use and land cover of Sangu-Matamuhuri watershed in the South-Eastern part of Bangladesh | |
| CN117437538A (en) | Tropical rainforest ecosystem space-time pattern feature extraction and prediction method |
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 |