CN113283281A - Zizania latifolia planting area extraction method based on multi-temporal remote sensing image - Google Patents
Zizania latifolia planting area extraction method based on multi-temporal remote sensing image Download PDFInfo
- Publication number
- CN113283281A CN113283281A CN202110220713.9A CN202110220713A CN113283281A CN 113283281 A CN113283281 A CN 113283281A CN 202110220713 A CN202110220713 A CN 202110220713A CN 113283281 A CN113283281 A CN 113283281A
- Authority
- CN
- China
- Prior art keywords
- months
- ndvi
- vegetation
- image
- band
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 21
- 244000085595 Zizania latifolia Species 0.000 title claims description 74
- 235000004259 Zizania latifolia Nutrition 0.000 title claims description 74
- 238000000034 method Methods 0.000 claims abstract description 46
- 230000003595 spectral effect Effects 0.000 claims abstract description 15
- 238000003066 decision tree Methods 0.000 claims abstract description 10
- 238000012795 verification Methods 0.000 claims abstract description 6
- 238000012545 processing Methods 0.000 claims abstract description 5
- 241000746966 Zizania Species 0.000 claims description 49
- 235000002636 Zizania aquatica Nutrition 0.000 claims description 47
- 241000209094 Oryza Species 0.000 claims description 22
- 235000007164 Oryza sativa Nutrition 0.000 claims description 22
- 235000009566 rice Nutrition 0.000 claims description 22
- 238000011160 research Methods 0.000 claims description 21
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 15
- 238000003306 harvesting Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 7
- 238000003973 irrigation Methods 0.000 claims description 7
- 230000002262 irrigation Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000000007 visual effect Effects 0.000 claims description 5
- 244000207740 Lemna minor Species 0.000 claims description 4
- 235000006439 Lemna minor Nutrition 0.000 claims description 4
- 235000001855 Portulaca oleracea Nutrition 0.000 claims description 4
- 238000013145 classification model Methods 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012512 characterization method Methods 0.000 claims 1
- 238000009826 distribution Methods 0.000 abstract description 13
- 238000001228 spectrum Methods 0.000 abstract description 6
- 239000011435 rock Substances 0.000 description 12
- 235000013399 edible fruits Nutrition 0.000 description 9
- 239000002689 soil Substances 0.000 description 7
- 238000009355 double cropping Methods 0.000 description 6
- 238000012544 monitoring process Methods 0.000 description 5
- 238000002360 preparation method Methods 0.000 description 5
- 238000002310 reflectometry Methods 0.000 description 5
- 235000013311 vegetables Nutrition 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000011835 investigation Methods 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 4
- 239000000463 material Substances 0.000 description 4
- 241000196324 Embryophyta Species 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000000630 rising effect Effects 0.000 description 3
- 238000005286 illumination Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 241001108921 Asclepias asperula Species 0.000 description 1
- 244000061508 Eriobotrya japonica Species 0.000 description 1
- 235000009008 Eriobotrya japonica Nutrition 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 241000209504 Poaceae Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 244000098338 Triticum aestivum Species 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000013339 cereals Nutrition 0.000 description 1
- 239000001679 citrus red 2 Substances 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 244000037666 field crops Species 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000009335 monocropping Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Abstract
The invention provides a method for extracting the planting area of cane shoots based on multi-temporal remote sensing images, which is characterized in that a sentinel 2 time sequence image covering the whole growth period of crops is used for processing and generating an NDVI time sequence spectrum, the cane shoots can be effectively distinguished from other ground objects by matching a decision tree classification method according to the growth characteristics of the cane shoots, the overall precision of crop area extraction is higher, and the decision tree classification result of combining the planting characteristics and the multi-temporal spectral index characteristics on the surface by result verification better conforms to the distribution condition of real ground objects.
Description
Technical Field
The invention belongs to the technical field of agricultural remote sensing, and particularly relates to a cane shoot planting area extraction method based on multi-temporal remote sensing images.
Background
Zizania latifolia (Zizania altifolia) gramineae and Zizania perennial shallow water herbaceous plants are one of water and eight immortals in China, and are mainly cultivated in Yangtze river basin in China and swamp and paddy field areas in south China. A set of more advanced and practical cultivation technology of the wild rice shoots in the yellow rock region in Taizhou city of Zhejiang province is gradually formed, the wild rice shoots in the yellow rock shed are white, tender and bright in skin color and excellent in commodity, and the wild rice shoots in the region become geographical sign registration products of agricultural products in China. The double-cropping water bamboo is a special industry in the yellow rock agriculture, the planting area is over 4500hm2 at present, and the double-cropping water bamboo has the largest facility water bamboo planting base in China. The method can timely and accurately acquire the planting area of the cane shoots, coordinate the planting area of vegetables and the planting structure of grain crops, provide reference for agricultural decision and provide basis for vegetable subsidy.
The water bamboos in the yellow rock area are generally planted in two seasons, the water bamboos harvested in summer are called as 'summer water bamboos', and the water bamboos harvested in autumn are called as 'autumn water bamboos'. The planting time of the summer wild rice shoots is about 11 months in the last year to 4 months in the next year, the summer wild rice shoots are generally harvested before and after the Qingming festival by adopting the production technology of covering cultivation with facility agricultural films and soil protection (see Zhejiang agricultural science 2016.57(10): page 1644 + 1646), meanwhile, the greenhouse is covered with films in about 12 th of the month, the films are generally uncovered in 3 rd of the month, and the agricultural films are completely uncovered after the basic Qingming (Changjiang vegetables 2009(16): page 102 + 103). The planting time of the autumn wild rice shoots is from 6 to 11 months, soil preparation is needed during the planting of the autumn wild rice shoots, irrigation is generally carried out in the field in 5 to 6 months after the summer wild rice shoots are harvested, then planting of the autumn wild rice shoots is carried out in the last ten days of 6 months to 7 months, the autumn wild rice shoots are harvested in 11 months, and more strong and strong tillering seedlings in autumn are reserved during the harvesting so as to increase the quality and yield of the early spring and summer wild rice shoots.
At present, the method for acquiring the area of the water bamboo mainly depends on the traditional manual survey method, a large amount of time is consumed, a large amount of manpower and material resources are needed, and the precision of the statistical survey is influenced by some inevitable subjective factors in the statistical process of the manual survey, such as statistical errors, inconsistent local standards, different measuring tools and the like. The remote sensing can carry out large-area synchronous monitoring, the timeliness is high, the remote sensing statistics of the planting area of the cane shoots is more accurate than that of a traditional method in the aspect of statistical results, the economic investment is low, and the method is not limited by regions.
At present, the research on the wild rice shoots is mostly focused on physiological and biochemical aspects such as cultivation measures, pest and disease occurrence and the like, the related research on the remote sensing extraction of the wild rice shoots is less, and the extracted crops are mainly focused on field crops such as wheat, rice, corn and the like.
The remote sensing classification method comprises two types of supervised classification and unsupervised classification. The supervision and classification is to select characteristic parameters according to samples provided by a known training area, use the trained characteristic parameters as decision rules, and classify the images to be classified according to the rules. The unsupervised classification is a method for classifying images according to known classification standards based on the class differences of different image ground objects in the feature space. Whole Jing and the like extract the distribution of the rice planting area in Dabie mountain areas by using an unsupervised classification method, and the method realizes the real-time dynamic monitoring of the rice planting area in a research area (see Chinese agronomy report, 2019.35(01): page 104-111). The Uulan Haote city of Sinkiang of Mongolian autonomous region within Mailisu is taken as a research area, and is compared with an unsupervised classification method based on three supervision classification methods such as a support vector machine method, a maximum likelihood method, an object-oriented classification method and the like, and research shows that the extraction accuracy is higher than that of the unsupervised classification method by adopting the supervision classification method (northern agriculture bulletin, 2019.47(05): page 119-126). The plum element and the like monitor the area of the rape by NDVI time sequence data and a decision tree classification method, and research shows that the identification result can effectively remove interference types such as non-vegetation, woodland and the like, and further extract the planting area and distribution of the rape (the science and science of Earth information, 2019.).
In addition, the characteristic aspect of crop extraction is mostly identified by using single characteristic or combination of multiple characteristics. The spectral characteristics of the rice canopy are analyzed by the Zhaoyang nations and the like, and the optimal time phase of remote sensing extraction of rice planting information is determined by comparison (modern agricultural science and technology, 2015(15): page 258-. However, data of other time phases are not utilized, and further research is needed compared with the rice extraction by a multi-time phase method. The time sequence remote sensing image not only has the spectrum information of a single time phase image, but also has a series of time information, has important significance in extracting crop distribution information, and a plurality of scholars use multi-time phase data to carry out remote sensing monitoring at present.
At present, the research of extracting the crop planting distribution in a large area mainly comprises the following aspects, namely the research of the remote sensing identification capability of crops under different remote sensing image conditions; secondly, the comparative study on the crop identification capability and precision is carried out based on different classification methods; and thirdly, combining the multi-temporal images and the multi-feature auxiliary crop classification. However, remote sensing identification for crops is mainly focused on a large number of crops, and related research on remote sensing identification of cane shoots is less. The crushing degree of the farmland plots in the yellow rock region is high, the varieties of crops are rich, and the phenomenon of 'same thing and different spectrum' of a plurality of crops exists, so that the extraction of the cane shoots is interfered. The acquisition of crop information in the past was based primarily on optical satellite data. For optical remote sensing, the distribution condition of crops is judged by analyzing the change of the spectral reflectivity of the current image, and the crops are extracted by constructing a vegetation index, so that the method has the limitation that the phenomenon of foreign matter co-spectrum exists between cane shoots and other crops in the same period, and each ground object cannot be well distinguished simply according to a single-time-phase image and spectral characteristics.
How to effectively and accurately identify the distribution and the area of the cane shoots to provide support for vegetable planting subsidy and production layout planning is a problem which is urgently needed to be solved in the field.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention aims to provide the double-cropping water bamboo planting area extraction method capable of accurately acquiring the area information and the spatial distribution of the water bamboo.
Specifically, the invention provides a cane shoot planting area extraction method based on a multi-temporal remote sensing image, which comprises the following steps:
s1, acquiring remote sensing multispectral images of main growth periods of zizania latifolia in a research area, extracting images containing red, green, blue and near-infrared bands, and then performing atmospheric correction processing on the images;
s2, constructing a decision tree classification model for classification according to the difference between the spectral index characteristics, the vegetation pattern characteristics and the color characteristics of the wild rice shoots and the confused crops;
s3, constructing a recognition model, distinguishing the cane shoots and estimating the area of the cane shoots.
In the above method, preferably, the step of S1, the obtaining of the remote sensing multispectral image of the main growth period of zizania latifolia in the research area is image data provided by a sentinel No. 2 satellite; more preferably, the central wavelength of the blue light band is 0.49 μm, and the resolution is better than 10 m; the central wavelength of the green light wave band is 0.56 μm, and the resolution is better than 10 m; the central wavelength of the red light wave band is 0.665 μm, and the resolution is better than 10 m; the central wavelength of the near infrared band is 0.842 μm, and the resolution is better than 10 m.
The water bamboos in the yellow rock area are generally planted in two seasons, the water bamboos harvested in summer are called as 'summer water bamboos', and the water bamboos harvested in autumn are called as 'autumn water bamboos'. S2, the identification basis of the vegetation pattern features comprises:
(1) covering a greenhouse agricultural film in a field before harvesting the summer zizania latifolia from 11 months in the last year to 3-4 months in the next year;
(2) after summer water bamboos are harvested and before autumn water bamboos are planted, a field irrigation stage is carried out;
(3) in the harvest stage of the autumn wild rice, the color characteristic is more green than that of the late rice harvested in the same period.
Further, according to the color characteristics of cane shoots, rice, fruit trees and forest lands in the sentinel image, the identification of the color characteristics in S2 includes:
visual interpretation is carried out by utilizing the characteristic that the wild rice stem plot in the month of 2 shows grey white;
extracting all plots which have water in the 6 months and are covered with the duckweed from the image of the 6 months;
and further screening out wild rice stem planting plots by using the index characteristic that the wild rice stems in 10 months are dark green in the field, and excluding the wild rice stem planting plots according to the image in 10 months.
Further, the spectral index features of S2 include:
calculating NDVI and NDWI vegetation indexes on the images from 2 months to 11 months;
the NDVI is a normalized vegetation index which is the ratio of the difference between the values of the near infrared band and the visible light red band and the sum of the values of the two bands; NDVI reflects the background effects of plant canopy and is related to vegetation coverage;
the formula for calculating the NDVI is as follows: NDVI ═ (NIR-R)/(NIR + R);
NDWI is a normalized ratio index based on the green band and the near infrared band; the effect of extracting the water body information in the image is good, the NDWI is based on the normalized ratio index of the green wave band and the near infrared wave band, and the calculation formula of the NDWI is as follows: NDWI ═ (G-NIR)/(G + NIR);
wherein NIR is an infrared band value, R is a visible light red band value, and G is a green band value.
Further, the constructing the recognition model in S3 includes:
s31, removing non-vegetation and artificial buildings by using NDVI;
s32, setting a threshold value by utilizing the NDWI and the NDVI to extract the water body range of 6 months;
s33, combining NDVI of 10 months to distinguish the cane shoots;
more preferably, the constructing the recognition model in S3 includes:
s31 extracting a 9-month image map, setting a threshold value NDVI >0.6, and when the condition is false, marking the image element as non-vegetation; when the condition is true, the pixel is marked as vegetation, and S32 judgment is carried out;
s32, extracting an image map in month 6, setting a threshold value of 0.4< NDVI <0.6, and marking the image element as other vegetation when the condition is false; when the condition is true, the step goes to the judgment of S33;
s33, extracting a 10-month image map, setting a threshold value NDVI >0.6, and marking the image element as other vegetation when the condition is false; when the condition is true, the picture element is marked as cane shoot.
Further, the method also comprises the step of evaluating the position precision by adopting a confusion matrix method; preferably, the evaluating the position accuracy by using the confusion matrix method comprises: and establishing 900 random verification points in a research area, visually interpreting each random point by combining a high-resolution image and a field situation, comparing with a remote sensing extraction result, making a confusion matrix, and respectively calculating the user precision, the drawing precision and the Kappa coefficient.
The invention provides a method for processing and generating an NDVI time sequence spectrum by using a remote sensing time sequence image covering the whole growth period of crops, which can effectively distinguish the water bamboo from other ground objects according to the growth characteristics of the water bamboo and by matching with a decision tree classification method, the overall accuracy of crop area extraction is higher, the verification result shows that the planting characteristics and the multi-temporal spectral index characteristics are combined, the decision tree classification result based on the two characteristics better conforms to the real ground object distribution condition, and the feasibility of remote sensing extraction of the water bamboo is verified.
The method provided by the invention provides a relevant basis for cane shoot investigation by applying high-resolution remote sensing data, and provides support for vegetable planting subsidy and production layout planning.
The decision tree model constructed by combining the multi-temporal spectral characteristics can effectively identify the planting area of the water bamboo, the user precision and the drawing precision of the water bamboo are respectively 96.00% and 94.00%, the Kappa coefficient is 0.93, and the user precision and the drawing precision of other ground features are respectively 95.00% and 93.00%.
Drawings
The technical solution of the present application is further explained below with reference to the drawings and the embodiments.
FIG. 1 is a map of the location of an area of investigation in an embodiment of the present application;
FIG. 2 is a diagram of a planting pattern of zizania latifolia in an embodiment of the present application;
FIG. 3 is a color characteristic diagram of double cropping water bamboo in the embodiment of the application;
FIG. 4 is a vegetation multi-temporal NDVI graph from month 2 to month 11 in the example of the application;
fig. 5 is a vegetation multi-temporal NDWI graph from month 2 to month 11 in the application example, wherein 5(a) is a respective feature multi-temporal NDWI graph, and 5(b) is a water bamboo multi-temporal NDWI graph;
FIG. 6 is a flow chart of the Zizania latifolia extraction technique.
FIG. 7 is the spatial distribution diagram of cane shoots in the example. Wherein, fig. 7(a) is the overall distribution of the cane in the scope of the embodiment, fig. 7(b) is an enlarged view of a local area in the scope of the embodiment, and fig. 7(c) is the planting condition of the cane in the local area.
FIG. 8 is a diagram of user accuracy and drawing accuracy of cane shoot.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Examples
The eastern part of yellow rock area of taizhou city in zhejiang, longitude and latitude 28 degrees 35 'N-28 degrees 45' N, 121 degrees 7 'E-121 degrees 12' E, located in the town of toutuo and the new front street are used as research areas (as shown in fig. 1), and the technical scheme and the beneficial effects of the invention are illustrated by using the image of the multi-view sentinel No. 2 from 2 months to 2020 within 11 months in 2020 by way of example.
The yellow rock area is in the middle of Taizhou city, the district is high in the west and low in the east, the west is more high mountains, the east has a part of plain, belongs to the Wenhuang plain, and the total area of the whole area is 988 square kilometers. The climate is mild and moist, the rainfall is abundant, the illumination is suitable for being clear in four seasons, the climate belongs to subtropical monsoon climate, the average temperature of the whole area for many years is 17 ℃, the day of continuous sunshine lasts for 247.9d, the average day of the year is 1955h, and the rainfall is distributed from west to east and decreases with the reduction of the altitude.
S1, acquiring remote sensing multispectral images of main growth periods of zizania latifolia in a research area, extracting images containing red, green, blue and near-infrared bands, and then performing atmospheric correction processing on the images;
the sentinel 2 carries a multispectral imager which can cover 13 spectral bands, the band range is from visible light, near infrared to short wave infrared, the sentinel is the only satellite containing 3 bands in the red side range, the highest spatial resolution is 10m, the width is 290km, and the resolution of the common working time of the two satellites can be improved to 5 d.
The data products provided by the sentinel No. 2 satellite are Level-1C and Level-2A, wherein the Level-1C is the apparent reflectivity data of the top of the atmospheric layer after radiation correction and geometric correction, and the Level-2A is the surface reflectivity data after atmospheric correction. In the embodiment, Level-2A data after atmospheric correction is selected, so that influences of factors such as cloud, atmosphere and illumination on reflection of ground objects are effectively eliminated, and an image time phase comprises a main growth period of the water bamboo, namely 10 scenes of images from 2 months in 2020 to 11 months in 2020. Four bands were selected, and the ranges and resolutions of each band are shown in table 1.
TABLE 1 sentinel No. 2 image band and resolution selected for this embodiment
| Wave band | Center wavelength (um) | Resolution (m) |
| Blue light | 0.490 | 10 |
| Green light | 0.560 | 10 |
| Red light | 0.665 | 10 |
| Near infrared | 0.842 | 10 |
S2, constructing a decision tree classification model for classification according to the difference between the spectral index characteristics, the vegetation pattern characteristics and the color characteristics of the wild rice shoots and the confused crops;
firstly, the planting pattern characteristics of the cane shoots are analyzed, and on the basis, the spectral index characteristics, the vegetation pattern characteristics, the color characteristics and the like of the cane shoots are analyzed, so that the basis for characteristic selection is provided for the extraction of the cane shoots; and then constructing a decision tree classification model for classification according to the difference between the spectral reflectivity characteristics and the vegetation index characteristics of the wild rice shoots and the confused crops.
Characteristic of planting pattern
The water bamboos in the yellow rock area are generally planted in two seasons, the water bamboos harvested in summer are called as 'summer water bamboos', and the water bamboos harvested in autumn are called as 'autumn water bamboos'. The planting time of the summer wild rice shoots is about 11 months in the last year to 4 months in the next year, the summer wild rice shoots are harvested before and after the clearing season by adopting a facility agricultural film covering cultivation and soil-up and protection production technology, meanwhile, the greenhouse film covering time is to cover a film in 12 months later, the film is uncovered in 3 months later, and the agricultural film is completely uncovered after the basic clearing. The planting time of the autumn wild rice shoots is from 6 to 11 months, soil preparation is needed during the planting of the autumn wild rice shoots, irrigation is generally carried out in the field in 5 to 6 months after the summer wild rice shoots are harvested, then planting of the autumn wild rice shoots is carried out in the last ten days of 6 months to 7 months, the autumn wild rice shoots are harvested in 11 months, and more strong and strong tillering seedlings in autumn are reserved during the harvesting so as to increase the quality and yield of the early spring and summer wild rice shoots.
The water bamboo planting process can be divided into a water bamboo planting stage and a soil preparation stage as shown in figure 2, and the following 3 points are summarized from the water bamboo planting mode and serve as the basis for visual interpretation and feature identification. The method comprises the following steps: firstly, covering a greenhouse agricultural film in a field before summer zizania latifolia is harvested from 11 months in the last year to 3-4 months in the next year; secondly, after the summer water bamboos are harvested, before the autumn water bamboos are planted, a field irrigation stage is carried out; and thirdly, compared with late rice harvested in the same period, the autumn water bamboo harvesting method is greener in color characteristic than the late rice.
Color characteristics
The main crops in the yellow rock area comprise rice and wild rice shoots, and the fruit trees mainly comprise orange trees and loquat trees. The yellow rock area rice is generally planted in the morning and evening by continuous cropping, and has the phenomenon of foreign matter co-spectrum with cane shoots in most periods. The color characteristics of fruit trees and forest lands do not change much all the year around. According to the color characteristics of the water bamboo, the rice, the fruit trees and the forest land on the sentinel image, the following results are obtained: the field blocks where the wild rice shoots are planted in the month 2 are covered by agricultural films, the field blocks are gray-colored in the image, the early rice is not planted in the month February, the field blocks are in a bare land state, the fruit trees and the forest lands are green lands all the year round, and the forest lands are greener than the fruit trees. Therefore, the characteristic that the wild rice stem land blocks in the month of 2 show grey color can be utilized for visual interpretation.
The summer cane shoots are harvested just in 6 months, the summer cane shoots are dark green on remote sensing images, and the color difference between fruit trees and forest lands is not large compared with that in February. The image of the 6 months is a key time phase of the identification of the water bamboo field block, all the field blocks which are watered in the 6 months are extracted by utilizing the characteristic that the field is covered with duckweed when the water bamboo field block is used for soil preparation and irrigation, and the extraction result at the moment comprises the field block planted with the water bamboo and other field blocks which are watered in the 6 months; and in 10 months, the wild rice shoots are dark green in the field, the field is obviously greener compared with late rice in a mature harvest period, at the moment, the wild rice shoots can be further screened out by utilizing the index characteristic of 10 months according to the result extracted in 6 months, and non-wild rice shoots are removed according to the image of 10 months.
Spectral index features
According to the multi-temporal sentinel No. 2 multispectral reflectivity image, visual interpretation is carried out by combining the Google image and the results of on-site actual investigation, 2103 sample points are selected in the research area, wherein 177 fruit tree sample points, 673 buildings, 790 water bamboo sample points, 289 woodlands and 174 rice sample points are selected, and the sample points are distributed as shown in figure 1. The buildings can be rejected with normalized vegetation indices.
Ndvi (normalized Difference Vegetation index) the normalized Vegetation index is defined as the ratio of the Difference between the values of the near infrared band (NIR) and the visible RED band (RED) and the sum of the values of these two bands. Can reflect the background influence of plant canopy and is related to vegetation coverage. NDWI (normalized Difference Water index) is based on the normalized ratio index of the green band and the near infrared band, and has better effect of extracting the water body information in the image.
A normalized vegetation index from month 2 to month 11 is calculated according to the vegetation index NDVI and NDWI calculation formula. According to the multi-time phase NDVI curve chart, the NDVI of various features is between 0.2 and 0.9, while the NDVI of the cane shoots is between 0.4 and 0.8. The zizania latifolia in the yellow rock area is generally planted in two seasons, the field with zizania latifolia all the year round is available, and even if no zizania latifolia is planted in the field, the NDVI of the zizania latifolia field is not low because duckweed is covered on water during soil preparation and irrigation. The NDVI of the wild rice stem plot is between 0.4 and 0.6 in the months from 2 to 7, the NDVI of the wild rice stem is gradually increased in the months from 7 to 11, the NDVI is generally stabilized between 0.7 and 0.8 in a harvest season, and the NDVI of the wild rice stem is higher than that of the rice in the same period in the month from 10. In fig. 4, NDVI of rice in 2 to 11 months shows a trend of falling-rising-falling as a whole, NDVI of 2 to 4 months falls, NDVI of 4 months falls to about 0.2, NDVI of 4 to 7 months shows a trend of rising first and then falling, late rice is planted in 8 to 11 months, and the trend of rising first and then falling is also shown. The forest land and the fruit tree have similar ascending and descending trends.
The NDWI is generally used for extracting water body information in an image, and it can be known from a multi-temporal NDWI graph 5a that the NDWI of each feature is generally between-0.8 and-0.2, the NDWI curve of the zizania latifolia generally shows a trend of rising first and then falling, the NDWI of the zizania latifolia is obviously different from other features in the 6 months to 7 months, as shown in fig. 5b, the NDWI of the zizania latifolia is greater than-0.4, and the other features are less than-0.4. The NDWI of rice showed a tendency of rising-falling-rising-falling overall.
S3, constructing an identification model, distinguishing water bamboo and estimating the area of the water bamboo:
the main crop in the research area is the wild rice shoots, and a small amount of rice is planted.
Through repeated comparison and tests, the NDVI can effectively remove non-vegetation and artificial buildings; the water body range of 6 months can be extracted by setting a threshold value by utilizing the NDWI sensitive to the water body and the NDVI sensitive to vegetation; the cane shoots can be distinguished by combining with NDVI of 10 months. Setting a threshold value NDVI to be more than 0.6 according to the NDVI of the vegetation in the 9 months, and screening out green vegetation; when the double cropping water bamboo is irrigated with water and prepared with soil in 6 months, the NDVI of the water bamboo plot is less than 0.6, and the NDWI is more than 0.4, so that other green vegetation can be removed, and the water body area can be extracted. FIG. 3 illustrates that the double cropping water bamboo has a significant difference in NDVI in month 10 from other vegetation, with NDVI >0.6 in month 10; when the condition is met, the classified material is water bamboo, otherwise, the classified material is other ground materials. The specific classification flow chart is shown in fig. 6. The classification result is shown in fig. 7, the zizania latifolia in fig. 7a is distributed more intensively, and is basically distributed in the region with flat terrain among the valleys, and the spatial distribution monitoring result is basically consistent with the field investigation condition.
Evaluation of model accuracy
In order to truly and objectively realize the evaluation of the model precision, the invention adopts a confusion matrix method to evaluate the position precision. And establishing 900 random verification points in a research area, visually interpreting each random point by combining a high-resolution image and a field situation, comparing with a remote sensing extraction result, making a confusion matrix, and respectively calculating the user precision, the drawing precision and the Kappa coefficient. The results are shown in FIG. 8. The Kappa coefficient of the wild rice stem is 0.93 through calculation, the user precision and the drawing precision of the wild rice stem are respectively 96.00% and 94.00%, the user precision and the drawing precision of other ground features are respectively 95.00% and 93.00%, and the error rate and the leakage rate are low. Based on two precision verification results of the area and the confusion matrix, the remote sensing extraction model of the water bamboo is verified to have high recognition precision, and remote sensing monitoring on the spatial distribution of the water bamboo in the area can be realized.
In light of the foregoing description of the preferred embodiments according to the present application, it is to be understood that various changes and modifications may be made without departing from the spirit and scope of the invention. The technical scope of the present application is not limited to the contents of the specification, and must be determined according to the scope of the claims.
Claims (10)
1. A method for extracting the planting area of cane shoots based on multi-temporal remote sensing images is characterized by comprising the following steps:
s1, acquiring remote sensing multispectral images of main growth periods of zizania latifolia in a research area, extracting images containing red, green, blue and near-infrared bands, and then performing atmospheric correction processing on the images;
s2, constructing a decision tree classification model for classification according to the difference between the spectral index characteristics, the vegetation pattern characteristics and the color characteristics of the wild rice shoots and the confused crops;
s3, constructing a recognition model, distinguishing the cane shoots and estimating the area of the cane shoots.
2. The method of claim 1, wherein the step of obtaining the remotely sensed multispectral image of the major growth period of zizania latifolia in the research area at S1 is performed by using influence data provided by a sentinel satellite 2.
3. The method of claim 2, wherein the blue band has a center wavelength of 0.49 μm and a resolution of better than 10 m; the central wavelength of the green light wave band is 0.56 μm, and the resolution is better than 10 m; the central wavelength of the red light wave band is 0.665 μm, and the resolution is better than 10 m; the central wavelength of the near infrared band is 0.842 μm, and the resolution is better than 10 m.
4. The method of claim 1, wherein the identification basis of the vegetation pattern features of S2 comprises:
(1) covering a greenhouse agricultural film in a field before harvesting the summer zizania latifolia from 11 months in the last year to 3-4 months in the next year;
(2) after summer water bamboos are harvested and before autumn water bamboos are planted, a field irrigation stage is carried out;
(3) in the harvest stage of the autumn wild rice, the color characteristic is more green than that of the late rice harvested in the same period.
5. The method of claim 1, wherein the identifying of the color feature of S2 comprises:
visual interpretation is carried out by utilizing the characteristic that the wild rice stem plot in the month of 2 shows grey white;
extracting all plots which have water in the 6 months and are covered with the duckweed from the image of the 6 months;
and further screening out wild rice stem planting plots by using the index characteristic that the wild rice stems in 10 months are dark green in the field, and excluding the wild rice stem planting plots according to the image in 10 months.
6. The method of claim 1, wherein the spectral index characterization of S2 comprises:
calculating NDVI and NDWI vegetation indexes on the images from 2 months to 11 months;
the NDVI is a normalized vegetation index which is the ratio of the difference between the values of the near infrared band and the visible light red band and the sum of the values of the two bands, and the calculation formula is as follows:
NDVI＝(NIR-R)/(NIR+R)；
NDWI is based on the normalized ratio index of the green band and the near-infrared band, and the calculation formula is as follows:
NDWI＝(G-NIR)/(G+NIR)；
wherein NIR is an infrared band value, R is a visible light red band value, and G is a green band value.
7. The method of claim 1, wherein the constructing of the recognition model at S3 includes:
s31, removing non-vegetation and artificial buildings by using NDVI;
s32, setting a threshold value by utilizing the NDWI and the NDVI to extract the water body range of 6 months;
s33, combining with NDVI of 10 months, water bamboo can be distinguished.
8. The method of claim 1, wherein the constructing of the recognition model at S3 includes:
s31 extracting a 9-month image map, setting a threshold value NDVI >0.6, and when the condition is false, marking the image element as non-vegetation; when the condition is true, the pixel is marked as vegetation, and S32 judgment is carried out;
s32, extracting an image map in month 6, setting a threshold value of 0.4< NDVI <0.6, and marking the image element as other vegetation when the condition is false; when the condition is true, the step goes to the judgment of S33;
s33, extracting a 10-month image map, setting a threshold value NDVI >0.6, and marking the image element as other vegetation when the condition is false; when the condition is true, the picture element is marked as cane shoot.
9. The method according to any one of claims 1 to 8, further comprising performing position accuracy evaluation by using a confusion matrix method.
10. The method of claim 9, wherein the evaluating the position accuracy by using the confusion matrix comprises: and establishing 900 random verification points in a research area, visually interpreting each random point by combining a high-resolution image and a field situation, comparing with a remote sensing extraction result, making a confusion matrix, and respectively calculating the user precision, the drawing precision and the Kappa coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110220713.9A CN113283281A (en) | 2021-02-26 | 2021-02-26 | Zizania latifolia planting area extraction method based on multi-temporal remote sensing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110220713.9A CN113283281A (en) | 2021-02-26 | 2021-02-26 | Zizania latifolia planting area extraction method based on multi-temporal remote sensing image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113283281A true CN113283281A (en) | 2021-08-20 |
Family
ID=77276200
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110220713.9A Pending CN113283281A (en) | 2021-02-26 | 2021-02-26 | Zizania latifolia planting area extraction method based on multi-temporal remote sensing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113283281A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117233123A (en) * | 2023-09-13 | 2023-12-15 | 宁波大学 | Large-scale remote sensing monitoring method and device for bacterial leaf blight of rice based on sentinel No. 2 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446875A (en) * | 2016-11-16 | 2017-02-22 | 航天恒星科技有限公司 | County remote-sensing scale-oriented crop planting area information extraction method and device |
| CN111079846A (en) * | 2019-12-20 | 2020-04-28 | 中国科学院遥感与数字地球研究所 | Apple identification method based on time series high-resolution remote sensing data |
| CN111462223A (en) * | 2020-04-22 | 2020-07-28 | 安徽大学 | Method for identifying soybean and corn planting area in Jianghuai region based on Sentinel-2 image |
| WO2020165671A1 (en) * | 2019-02-11 | 2020-08-20 | Università Degli Studi Di Palermo | Method for monitoring vegetation ground covers |
| CN111860149A (en) * | 2020-06-11 | 2020-10-30 | 中科禾信遥感科技（苏州）有限公司 | Remote sensing identification method and device for overwintering rape and wheat |
-
2021
- 2021-02-26 CN CN202110220713.9A patent/CN113283281A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446875A (en) * | 2016-11-16 | 2017-02-22 | 航天恒星科技有限公司 | County remote-sensing scale-oriented crop planting area information extraction method and device |
| WO2020165671A1 (en) * | 2019-02-11 | 2020-08-20 | Università Degli Studi Di Palermo | Method for monitoring vegetation ground covers |
| CN111079846A (en) * | 2019-12-20 | 2020-04-28 | 中国科学院遥感与数字地球研究所 | Apple identification method based on time series high-resolution remote sensing data |
| CN111462223A (en) * | 2020-04-22 | 2020-07-28 | 安徽大学 | Method for identifying soybean and corn planting area in Jianghuai region based on Sentinel-2 image |
| CN111860149A (en) * | 2020-06-11 | 2020-10-30 | 中科禾信遥感科技（苏州）有限公司 | Remote sensing identification method and device for overwintering rape and wheat |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117233123A (en) * | 2023-09-13 | 2023-12-15 | 宁波大学 | Large-scale remote sensing monitoring method and device for bacterial leaf blight of rice based on sentinel No. 2 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106372592B (en) | A kind of winter wheat planting area calculation method based on winter wheat area index | |
| Tennakoon et al. | Estimation of cropped area and grain yield of rice using remote sensing data | |
| Usha et al. | Potential applications of remote sensing in horticulture—A review | |
| CN107273820B (en) | Remote sensing classification method and system for land coverage | |
| CN114821362B (en) | Multi-source data-based rice planting area extraction method | |
| CN109740570B (en) | Terrace remote sensing identification method based on terrain index and climate difference | |
| Kodani et al. | Seasonal patterns of canopy structure, biochemistry and spectral reflectance in a broad-leaved deciduous Fagus crenata canopy | |
| Yang et al. | Airborne multispectral imagery for mapping variable growing conditions and yields of cotton, grain sorghum, and corn | |
| US20220392215A1 (en) | System and Method for Mapping Land Cover Types with Landsat, Sentinel-1, and Sentinel-2 Images | |
| CN111209871A (en) | Rape planting land remote sensing automatic identification method based on optical satellite image | |
| CN116129276A (en) | Remote sensing fine classification method for main grain crops in terrain complex region | |
| CN110598514B (en) | Method for monitoring plot scale crop seeding area of land reclamation project area | |
| Pauly | Applying conventional vegetation vigor indices to UAS-derived orthomosaics: issues and considerations | |
| CN111079846A (en) | Apple identification method based on time series high-resolution remote sensing data | |
| Pieri et al. | Estimation of actual evapotranspiration in fragmented Mediterranean areas by the spatio-temporal fusion of NDVI data | |
| CN113283281A (en) | Zizania latifolia planting area extraction method based on multi-temporal remote sensing image | |
| Abdelraouf et al. | Comparative analysis of some winter crops area estimation using landsat-8 and sentinal-2 satellite imagery | |
| Chea et al. | Sugarcane canopy detection using high spatial resolution UAS images and digital surface model. | |
| Machado et al. | Stress conditions in soybean areas based on measurements of soil-plant-atmosphere system and UAV images | |
| Jovanović et al. | Crop yield estimation in 2014 for Vojvodina using methods of remote sensing | |
| CN114332628B (en) | Ginger rapid remote sensing extraction method based on typical physical condition and film network characteristics | |
| Safdary et al. | Application of landscape metrics and object-oriented remote sensing to detect the spatial arrangement of agricultural land | |
| Yoshikawa et al. | Estimating variable acreage of cultivated paddy fields from preceding precipitation in a tropical watershed utilizing Landsat TM/ETM | |
| El-Sharkawy | Precision agriculture using advanced remote sensing techniques for peanut crop in Arid Land | |
| Sbahi et al. | Evaluation of the Efficiency of Agricultural Production in the Pivotal Farms Utilizing Remote Sensing Techniques |
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 |