CN109446962B - Land cover annual change detection method and system - Google Patents
Land cover annual change detection method and system Download PDFInfo
- Publication number
- CN109446962B CN109446962B CN201811222022.7A CN201811222022A CN109446962B CN 109446962 B CN109446962 B CN 109446962B CN 201811222022 A CN201811222022 A CN 201811222022A CN 109446962 B CN109446962 B CN 109446962B
- Authority
- CN
- China
- Prior art keywords
- time sequence
- ndvi
- target
- target year
- remote sensing
- 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.)
- Expired - Fee Related
Links
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
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Geophysics And Detection Of Objects (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the invention provides a method and a system for detecting change of land cover annual ring.A NDVI time sequence data of a target area in a first target year and a NDVI time sequence data of the target area in a second target year are respectively determined through time sequence remote sensing data of the target area in the first target year and the second target year; on the basis that the NDVI time sequence data of two different years are reconstructed by the preset remote sensing time sequence model, similarity comparison of time sequence curves of the two different years is carried out by using a DTW algorithm, and an annual change detection result of land coverage is obtained. The embodiment of the invention fully excavates the available value of the time sequence remote sensing data, and has great application prospect in the aspect of using the remote sensing time sequence image to change and detect. And the problem of displacement on a time axis can be solved to a certain extent through a DTW algorithm, the influence of abnormal values such as noise, cloud and the like can be reduced, a better matching effect is obtained, and the detection precision of the change of the land cover annual period is improved.
Description
Technical Field
The embodiment of the invention relates to the technical field of remote sensing image change detection, in particular to a land cover annual change detection method and a land cover annual change detection system.
Background
The land cover change is the comprehensive reaction of a plurality of elements on the earth surface, has direct influence on the energy balance, climate change, water circulation and the like on the earth surface, and is the focus of the attention of a plurality of disciplines. The remote sensing data is an important data source for identifying the land cover change due to the macroscopicity, the periodicity and the continuity of the remote sensing data. The time sequence remote sensing image can provide a record of a long-time sequence of the earth surface state, reflect the dynamic change condition of the earth surface in a long-time range, and provide a reliable data source for the space-time identification of land coverage change. The method for fully mining the temporal dimensional information contained in the remote sensing time sequence data set and developing the automatic detection method of the land cover change based on the time sequence remote sensing is one of the hot spots of the research in the field of remote sensing technology in recent years.
In the process of using the time series remote sensing data to detect land cover change, singular values formed by influences of clouds, cloud shadows, fog, haze and the like in the remote sensing data need to be eliminated firstly, namely, noise of a time series curve is eliminated, and the process is a process of filtering the remote sensing time series data and is also a process of constructing a time series model. In order to realize automatic detection of land cover change, multi-temporal remote sensing data is used as an information source, and a plurality of methods such as an image difference method, a spectral feature variation method, a principal component analysis method, a false color synthesis method, a waveband replacement method, a comparison method after classification, a multiband cross correlation analysis method, a change vector analysis method and the like are developed, and have advantages, however, land cover change detection still has difficulties: firstly, the phenomena of 'same object, different spectrum and same spectrum foreign matter' cause that the earth surface change and the image change characteristics can not be in one-to-one correspondence; secondly, the land cover change comprises intra-class change and inter-class change, namely the land cover change caused by the same kind of vegetation and the land cover change caused by the vegetation type change, the change degree also comprises partial change and all change, and how to define the change is also a difficulty aiming at different research targets; third, instability in land cover changes makes change detection more complex.
The key of the change detection by utilizing the time characteristics of the land coverage in the remote sensing images of the long-time sequence is to compare the similarity analysis of the two time sequences, namely, to quantitatively evaluate the similarity between the time sequence curve of the pixel to be detected and the time sequence curve of the target reference pixel. The Euclidean distance method is generally adopted in the time sequence similarity calculation method, the Euclidean distance is used for calculating the real distance between two points at the same time, and in the process of detecting land coverage change, due to the reasons of surface temperature, crop growth, land feature change and the like, time sequence curves of land features stretch out and draw back and deviate on an axis in different years, so that the problem of displacement of the time sequence curves on the time axis cannot be solved by adopting the Euclidean distance method.
Disclosure of Invention
To overcome the above problems or at least partially solve the above problems, embodiments of the present invention provide a method and a system for detecting an annual change in the soil coverage.
In a first aspect, an embodiment of the present invention provides a method for detecting an annual change in land coverage, including:
respectively determining normalized differential vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on the time sequence remote sensing data of the target area in the first target year and the second target year;
respectively fitting NDVI time sequence data of the target region in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target region in the first target year and a second NDVI time sequence curve of the target region in the second target year;
determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determining an annual change detection result of the target region in the first target year and the second target year based on the DTW distance value.
In a second aspect, an embodiment of the present invention provides a system for detecting an annual change in land cover, including: the NDVI time sequence data acquisition module, the NDVI time sequence curve acquisition module and the annual change detection result acquisition module. Wherein the content of the first and second substances,
the NDVI time sequence data acquisition module is used for respectively determining normalized difference vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on time sequence remote sensing data of the target area in the first target year and the second target year;
the NDVI time sequence curve acquisition module is used for respectively fitting the NDVI time sequence data of the target area in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target area in the first target year and a second NDVI time sequence curve of the target area in the second target year;
the annual change detection result acquisition module is used for determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm and determining annual change detection results of the target area in the first target year and the second target year based on the DTW distance value.
In a third aspect, an embodiment of the present invention provides an electronic device, including:
at least one processor, at least one memory, a communication interface, and a bus; wherein the content of the first and second substances,
the processor, the memory and the communication interface complete mutual communication through the bus;
the memory stores program instructions executable by the processor, which invokes the program instructions to perform the method of land cover annual change detection provided by the first aspect.
In a fourth aspect, embodiments of the present invention provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method of detecting annual change in land cover provided in the first aspect.
According to the method and the system for detecting the change of the land cover annual period, the NDVI time sequence data of the target area in the first target year and the NDVI time sequence data of the target area in the second target year are respectively determined through the time sequence remote sensing data of the target area in the first target year and the second target year; on the basis that the NDVI time sequence data of two different years are reconstructed by the preset remote sensing time sequence model, similarity comparison of time sequence curves of the two different years is carried out by using a DTW algorithm, and an annual change detection result of land coverage is obtained. The embodiment of the invention fully excavates the available value of the time sequence remote sensing data, and has great application prospect in the aspect of using the remote sensing time sequence image to change and detect. And the problem of displacement on a time axis can be solved to a certain extent through a DTW algorithm, the influence of abnormal values such as noise, cloud and the like can be reduced, a better matching effect is obtained, and the detection precision of the change of the land cover annual period is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a land cover annual change detection method provided in an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a land cover annual change detection system provided in an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an electronic device provided in an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the embodiments of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience in describing the embodiments of the present invention and simplifying the description, but do not indicate or imply that the referred devices or elements must have specific orientations, be configured in specific orientations, and operate, and thus, should not be construed as limiting the embodiments of the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the embodiments of the present invention, it should be noted that, unless explicitly stated or limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. Specific meanings of the above terms in the embodiments of the present invention can be understood in specific cases by those of ordinary skill in the art.
Fig. 1 is a schematic flow chart of a method for detecting an annual change in land cover provided in an embodiment of the present invention, and as shown in fig. 1, the method for detecting an annual change in land cover provided in an embodiment of the present invention specifically includes:
s1, respectively determining normalized differential vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on time sequence remote sensing data of the target area in the first target year and the second target year;
s2, respectively fitting the NDVI time sequence data of the target area in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target area in the first target year and a second NDVI time sequence curve of the target area in the second target year;
s3, determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determining the annual change detection result of the target area in the first target year and the second target year based on the DTW distance value.
Specifically, the embodiment of the invention aims to determine whether the land cover changes between two different years according to the obtained time series remote sensing data of the target area in the two different years. For example, when the first target year is 2008 and the second target year is 2009, the time-series remote sensing data of the target area in 2008 and the time-series remote sensing data of the target area in 2009 are acquired first. The time-series remote sensing data refers to remote sensing data arranged in time sequence, and specifically refers to the remote sensing data of each day in a first target year (2008) and arranged in time sequence, and the remote sensing data of each day in a second target year (2009) and arranged in time sequence. The method for acquiring the time sequence remote sensing data in the embodiment of the invention can be determined according to needs, and Landsat satellites can be selected as a preferred scheme for acquisition. The remote sensing data refers to spectral data, namely the reflectivity of the earth surface at each acquired wave band.
Because the NDVI can eliminate partial radiation errors and noises and expand the difference between different object categories, the NDVI time sequence data is selected for research. According to the time sequence remote sensing data of the target area in the first target year, the Normalized Difference Vegetation Index (NDVI) time sequence data of the target area in the first target year can be determined; similarly, according to the time sequence remote sensing data of the target area in the second target year, the NDVI time sequence data of the target area in the second target year can be determined. The implementation scheme for determining the NDVI time sequence data according to the time sequence remote sensing data can be realized by a calculation formula of the NDVI, the NDVI is equal to the difference value of the remote sensing data of the near infrared band and the remote sensing data of the red light band, the difference value is divided by the sum of the remote sensing data of the two bands, and the formula is specifically expressed as follows:
wherein R isNIRRepresenting the surface reflectivity in the near-infrared band, i.e. remote sensing data, R, in the near-infrared bandRedAnd the remote sensing data represents the surface reflectivity of the red light wave band, namely the remote sensing data of the red light wave band.
For different types of land coverage, the value of NDVI is between-1 and 1: for water bodies, the value of NDVI is usually negative; for bare land and rock, the value of NDVI is usually around 0; for farmlands and woodlands with high vegetation coverage, the value of NDVI is increased along with the increase of the vegetation coverage.
According to the embodiment of the invention, after the NDVI time sequence data of the target area in the first target year and the second target year are respectively determined, the NDVI time sequence data of the target area in the first target year and the second target year are respectively fitted based on the preset remote sensing time sequence model, and a first NDVI time sequence curve of the target area in the first target year and a second NDVI time sequence curve of the target area in the second target year are determined. The preset remote sensing time sequence model in the embodiment of the invention refers to a continuous function model which can fit discrete NDVI time sequence data. And fitting the NDVI time sequence data of the first target year and the NDVI time sequence data of the second target year through a preset remote sensing time sequence model, and respectively determining the parameter value of the preset remote sensing time sequence model corresponding to the NDVI time sequence data of the first target year and the parameter value of the preset remote sensing time sequence model corresponding to the NDVI time sequence data of the second target year to obtain a first NDVI time sequence curve and a second NDVI time sequence curve.
After the first NDVI timing curve and the second NDVI timing curve are determined, based on a Dynamic Time Warping (DTW) algorithm, determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve, and determining an annual change detection result of the target area in the first target year and the second target year based on the DTW distance value. The DTW algorithm is based on the idea of Dynamic Programming (DP), and is used to describe a one-to-one correspondence relationship between two input sequences in time when a time-warping function satisfies a certain condition, so that the minimum accumulated distance is ensured when the two sequences are matched. And determining the DTW distance value, namely determining the similarity between the NDVI time sequence curves of the two target years according to the DTW distance value, and further determining the annual change detection results of the target area in the first target year and the second target year. The detection result of the annual variation in the embodiment of the present invention generally includes two types: one is a change, i.e. the land cover changes in a first target year and a second target year; one is no change, i.e. the land cover in the first target year and the second target year is not changed. The lower the similarity, the change is indicated, and the higher the similarity, the no change is indicated.
According to the land cover annual change detection method provided by the embodiment of the invention, NDVI time sequence data of the target area in the first target year and the second target year are respectively determined through time sequence remote sensing data of the target area in the first target year and the second target year; on the basis that the NDVI time sequence data of two different years are reconstructed by the preset remote sensing time sequence model, similarity comparison of time sequence curves of the two different years is carried out by using a DTW algorithm, and an annual change detection result of land coverage is obtained. The embodiment of the invention fully excavates the available value of the time sequence remote sensing data, and has great application prospect in the aspect of using the remote sensing time sequence image to change and detect. And the problem of displacement on a time axis can be solved to a certain extent through a DTW algorithm, the influence of abnormal values such as noise, cloud and the like can be reduced, a better matching effect is obtained, and the detection precision of the change of the land cover annual period is improved.
On the basis of the above embodiment, in the land cover annual variation detection method provided in the embodiment of the present invention, the preset remote sensing time sequence model is a function model determined by a sine term, a cosine term, a constant term, a primary term, and a secondary term.
Specifically, in the prior art, many time series reconstruction algorithms appeared in the process of applying remote sensing time series data, which mainly include: and the maximum value synthesis method is used for taking the value of each point in the time sequence as the value of a new image corresponding to the position of the point according to the magnitude sequence in a specified time period. The method can eliminate the interference of cloud and atmosphere to a certain extent, but has simple processing, does not fully consider the influence of surface bidirectional reflection, and loses a lot of useful information. The optimal exponential slope extraction method considers that vegetation change is stable, and when a sudden change appears in a time sequence, the sudden change is probably caused by cloud or a sensor visual angle, so a sliding time window is adopted to judge a domain value and identify noise in the time sequence. The method has no influence on the gradual change time sequence, only influences the part which suddenly descends and then gradually ascends in the time sequence curve, cannot effectively remove the higher time sequence value caused by the atmospheric condition, and the setting of the sliding window threshold value also needs to be tested repeatedly. The mean iterative filtering method is one filtering and smoothing method to eliminate high frequency change in time sequence data. The algorithm has good smoothing effect and is simple, but important change information in a time sequence is ignored, and the iterative operation consumes a large amount of time. Savitzky-Golay filtering (S-G filtering), a least squares based convolution algorithm that is not constrained by the time-series time scale, the data space, and the sensors, and empirically derives the polynomial fitting order and the filter interval length. Generally, the larger the interval length of the filter is, the smoother the obtained result is, and thus, important change information in the time sequence is ignored. The Fourier transform method is a process of synthesizing a complex time sequence curve by a series of sine and cosine wave superposition modes to realize time sequence filtering. The curve obtained by Fourier transform has obvious vegetation growth change characteristics and is smooth, but because the symmetry of a fitting formula is strict, the method is difficult to realize on the problem of asymmetric information extraction, such as changes caused by agricultural activities, residential area construction or other artificial influences, and cannot be completely applied to remote sensing time sequence data filtering and reconstruction. Therefore, in the embodiment of the invention, the preset remote sensing time sequence model is adopted to realize the fitting of the NDVI time sequence data of the target area in the first target year and the second target year, so as to determine the first NDVI time sequence curve of the target area in the first target year and the second NDVI time sequence curve in the second target year. The preset remote sensing time sequence model adopted in the embodiment of the invention is a remote sensing time sequence model constructed in advance by an improved least square method, and particularly is a function model determined by a sine term, a cosine term, a constant term, a primary term and a secondary term. By adopting the preset remote sensing time sequence model, not only can obvious vegetation growth change characteristics be fitted, but also a good fitting effect can be obtained in an area with frequent human activities and complicated and changeable land coverage types.
On the basis of the above embodiment, in the method for detecting change of land cover between years provided in the embodiment of the present invention, the specific representation form of the preset remote sensing time sequence model is as follows:
wherein, a0Is a constant term of1、a2、a3And a4The coefficient of cosine term, the coefficient of sine term, the coefficient of primary term and the coefficient of secondary term are respectively constant; x is a time parameter in days in the first or second target year, and T is the number of days in the first or second target year;
NDVI timing data corresponding to x.
Preferably, x in the embodiment of the present invention is the number of days of julian days, i.e., the number of days of a year. Coefficient of cosine term a11, justCoefficient of chord term a2Expresses the annual variation, the first term coefficient a3Expresses the annual variation and the coefficient of the quadratic term a4An annual perturbation is expressed.
On the basis of the foregoing embodiment, the method for detecting an annual change in land cover provided in the embodiment of the present invention, before determining NDVI time series data of the target area in the first target year and the second target year, further includes:
respectively preprocessing the time sequence remote sensing data of the target area in the first target year and the second target year;
the pre-treatment at least comprises: radiation calibration, atmospheric correction and geometric correction.
Specifically, in the embodiment of the present invention, because the obtained time series remote sensing data of the target area in the first target year and the second target year are directly used, the obtained detection result of the change between years may be inaccurate, and therefore, the obtained time series remote sensing data needs to be preprocessed, where the preprocessing operation at least includes radiometric calibration, atmospheric correction, and geometric correction. The radiation calibration and the atmospheric correction are two processes of radiation correction, and the radiation calibration converts a remote sensing image pixel brightness value (DN) of remote sensing data into a radiance value of the top of an atmospheric layer; atmospheric correction is to eliminate the influence of factors such as atmosphere and illumination on the reflection of ground objects. Geometric correction is to eliminate the geometric distortion of the image, improve the geographic positioning precision and match the position of the actual ground object. The radiometric calibration is realized by Processing through a radiometric calibration module in ENVI remote sensing image professional Processing software, the atmospheric correction is realized by Processing through an atmospheric correction tool in a Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS), and the geometric correction is realized by Processing through a geometric correction model in the ENVI remote sensing image professional Processing software, so that the error of a geometric position is controlled within one pixel.
It should be noted that the preprocessing operation in the embodiment of the present invention may further include image clipping, and the image clipping may be performed on a specific area to obtain a remote sensing time sequence image of the target area after the radiation calibration, the atmospheric correction, and the geometric correction.
On the basis of the foregoing embodiment, the method for detecting an annual change in land cover provided in the embodiment of the present invention is configured to determine a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a dynamic time warping DTW algorithm, and specifically includes:
determining the shortest normalization path of the first NDVI timing curve and the second NDVI timing curve through the DTW algorithm according to the first NDVI timing curve and the second NDVI timing curve, and calculating the length value of the shortest normalization path;
and calculating the DTW distance value based on the length value.
Specifically, in the method for detecting an annual change in land cover provided in the embodiment of the present invention, it is assumed that two time series for calculating the similarity are: the first NDVI timing curve is X ═ X1,x2,x3…xm}，x1、x2、x3…xmM discrete time coordinates, x, on the first NDVI timing curve1Is the starting point, x, of the first NDVI timing curvemThe end point of the first NDVI timing curve, the length of the first NDVI timing curve is | X |; the second NDVI timing curve is Y ═ Y1,y2,y3…yn}，y1、y2、y3…ynN discrete time coordinates, y, respectively on the second NDVI timing curve1Is the starting point of the second NDVI timing curve, ynThe second NDVI timing curve has a length | Y | at the end of the second NDVI timing curve. The form of the rounding path may be expressed as: w ═ W1,w2,w3,...,wk,...,wKAnd (c) in which Max (| X |, | Y |) is less than or equal to | W |, is less than or equal to | X | + | Y |. w is akIs in the form of wk(i, j), wherein i represents the ith time coordinate in X and j represents the jth time coordinate in Y. The normalization path W starts at (1,1) and ends at (m, n), ensuring that each time coordinate in X and Y occurs in W, W being the normalized pathkThe subscript of (1) is monotonically increasing, i.e. wk＝(i,j)，wk+1I ≦ i '≦ i +1, j ≦ j' ≦ j + 1. And traversing each time coordinate in the first NDVI time sequence curve and the second NDVI time sequence curve to finally obtain a shortest regular path. And calculating the length value of the shortest reduction path, namely the sum of the distances of all path sections in the shortest reduction path, and calculating the DTW distance value according to the length value of the shortest reduction path.
In the embodiment of the invention, the shortest rounding path is determined, and the DTW distance value is calculated according to the length value of the shortest rounding path, so that an implementable scheme is provided for obtaining the DTW distance value.
On the basis of the above embodiment, in the method for detecting an annual change in land cover provided in the embodiment of the present invention, the length value of the shortest regression route is calculated by the following formula:
D(i,j)＝Dist(i,j)+min{D(i-1,j),D(i,j-1),D(i-1,j-1)}
wherein i is the ith time coordinate in the first NDVI time sequence curve, j is the jth time coordinate in the second NDVI time sequence curve, i is not less than 1, j is not less than 1, Dist (i, j) ═ xi-yjAnd has D (1,1) ═ Dist (1,1), xiIs the ith time coordinate value in the first NDVI timing curve,yjis the jth time coordinate value in the second NDVI timing curve.
Specifically, in the embodiment of the present invention, the length value of the shortest reduction path is calculated through the above formula, and a feasible scheme is provided for determining the length value of the shortest reduction path. It should be noted that, in the embodiment of the present invention, the process of calculating the length value of the shortest rounding path is also a process of determining the shortest rounding path, and the two processes supplement each other.
On the basis of the foregoing embodiment, in the method for detecting an annual change in land cover provided in the embodiment of the present invention, after determining the length value of the shortest normalized route, the DTW distance value needs to be calculated according to the length value, which is specifically determined by the following formula:
DDTW＝D/(m+n)
wherein D isDTWIs DTW distance value, D is the finally determined shortest regressionLength value of the whole path.
On the basis of the foregoing embodiment, the method for detecting an annual change in land cover provided in the embodiment of the present invention specifically includes, based on the DTW distance value, determining an annual change detection result of the target area in the first target year and the second target year:
comparing the DTW distance value to a change detection threshold;
and if the DTW distance value is judged to be larger than the change detection threshold value, determining that the annual change detection result is changed.
Specifically, as the DTW distance value is larger, the NDVI timing curves representing two years are more similar, i.e., the land cover change is larger for two years. Therefore, in the embodiment of the invention, the DTW distance values of each land cover type in two years are comprehensively considered, the size of the change detection threshold is set to be 1.0, the change conditions of all land cover types can be extracted, and the annual change detection result of the land cover is finally obtained. In the embodiment of the invention, after the DTW distance value is obtained, the DTW distance value is compared with the change detection threshold, if the DTW distance value is judged to be larger than the change detection threshold, the annual change detection result is determined to be changed, otherwise, the annual change detection result is determined not to be changed.
According to the embodiment of the invention, the change detection threshold is introduced to determine the annual change detection result, so that the annual change detection result is more visual.
As shown in fig. 2, on the basis of the above embodiment, an embodiment of the present invention provides a system for detecting an annual change in soil coverage, including: the NDVI time sequence data acquisition module 21, the NDVI time sequence curve acquisition module 22 and the annual change detection result acquisition module 23. Wherein the content of the first and second substances,
the NDVI time series data acquisition module 21 is configured to determine normalized difference vegetation index NDVI time series data of a target area in a first target year and a second target year respectively based on time series remote sensing data of the target area in the first target year and the second target year;
the NDVI timing curve obtaining module 22 is configured to respectively fit the NDVI timing data of the target region in the first target year and the second target year based on a preset remote sensing timing model, and determine a first NDVI timing curve of the target region in the first target year and a second NDVI timing curve of the target region in the second target year;
the annual change detection result obtaining module 23 is configured to determine a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determine an annual change detection result of the target area in the first target year and the second target year based on the DTW distance value.
Specifically, the NDVI timing data acquiring module 21 may determine Normalized Difference Vegetation Index (NDVI) timing data of the target region in the first target year according to the timing remote sensing data of the target region in the first target year; similarly, according to the time sequence remote sensing data of the target area in the second target year, the NDVI time sequence data of the target area in the second target year can be determined. The implementation scheme for determining the NDVI time sequence data according to the time sequence remote sensing data can be realized by a calculation formula of the NDVI, the NDVI is equal to the difference value of the remote sensing data of the near infrared band and the remote sensing data of the red light band, the difference value is divided by the sum of the remote sensing data of the two bands, and the formula is specifically expressed as follows:
wherein R isNIRRepresenting the surface reflectivity in the near-infrared band, i.e. remote sensing data, R, in the near-infrared bandRedAnd the remote sensing data represents the surface reflectivity of the red light wave band, namely the remote sensing data of the red light wave band.
After the NDVI timing data of the target area in the first target year and the second target year are respectively determined by the NDVI timing data acquisition module 21, the NDVI timing curve acquisition module 22 respectively fits the NDVI timing data of the target area in the first target year and the second target year based on the preset remote sensing timing model, and determines a first NDVI timing curve of the target area in the first target year and a second NDVI timing curve of the target area in the second target year. The preset remote sensing time sequence model in the embodiment of the invention refers to a continuous function model which can fit discrete NDVI time sequence data. And fitting the NDVI time sequence data of the first target year and the NDVI time sequence data of the second target year through a preset remote sensing time sequence model, and respectively determining the parameter value of the preset remote sensing time sequence model corresponding to the NDVI time sequence data of the first target year and the parameter value of the preset remote sensing time sequence model corresponding to the NDVI time sequence data of the second target year to obtain a first NDVI time sequence curve and a second NDVI time sequence curve.
After the NDVI timing curve obtaining module 22 determines the first NDVI timing curve and the second NDVI timing curve, the annual change detection result obtaining module 23 determines a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determines the annual change detection results of the target area in the first target year and the second target year based on the DTW distance value.
The functions of the modules in the land cover annual change detection system provided in the embodiment of the present invention correspond to the operation flows of the method embodiments one to one, and the details are not repeated herein in the embodiment of the present invention.
The land cover annual change detection system provided by the embodiment of the invention respectively determines the NDVI time sequence data of the target area in the first target year and the second target year through the time sequence remote sensing data of the target area in the first target year and the second target year; on the basis that the NDVI time sequence data of two different years are reconstructed by the preset remote sensing time sequence model, similarity comparison of time sequence curves of the two different years is carried out by using a DTW algorithm, and an annual change detection result of land coverage is obtained. The embodiment of the invention fully excavates the available value of the time sequence remote sensing data, and has great application prospect in the aspect of using the remote sensing time sequence image to change and detect. And the problem of displacement on a time axis can be solved to a certain extent through a DTW algorithm, the influence of abnormal values such as noise, cloud and the like can be reduced, a better matching effect is obtained, and the detection precision of the change of the land cover annual period is improved.
On the basis of the above embodiment, the system for detecting change of soil coverage between years provided in the embodiment of the present invention further includes a preprocessing module, where the preprocessing module is configured to: respectively preprocessing time sequence remote sensing data of the target area in the first target year and the second target year before determining the NDVI time sequence data of the target area in the first target year and the second target year; the pre-treatment at least comprises: radiation calibration, atmospheric correction and geometric correction.
On the basis of the foregoing embodiment, in the system for detecting change of land cover annual ring provided in the embodiment of the present invention, the annual ring change detection result obtaining module 23 includes a DTW distance value operator module, and the DTW distance value operator module is configured to: determining the shortest normalization path of the first NDVI time sequence curve and the second NDVI time sequence curve through the DTW algorithm according to the first NDVI time sequence curve and the second NDVI time sequence curve, and calculating the length value of the shortest normalization path; and calculating the DTW distance value based on the length value.
On the basis of the foregoing embodiment, in the system for detecting change of soil coverage annual ring provided in the embodiment of the present invention, the annual change detection result obtaining module 23 further includes an annual change detection result determining sub-module, and the annual change detection result determining sub-module is configured to: comparing the DTW distance value to a change detection threshold; and if the DTW distance value is judged to be larger than the change detection threshold value, determining that the annual change detection result is changed.
As shown in fig. 3, on the basis of the above embodiment, an embodiment of the present invention further provides an electronic device, including: a processor (processor)301, a memory (memory)302, a communication Interface (Communications Interface)303, and a bus 304; wherein the content of the first and second substances,
the processor 301, the memory 302 and the communication interface 303 complete communication with each other through the bus 304. The memory 302 stores program instructions executable by the processor 301, and the processor 301 is configured to call the program instructions in the memory 302 to perform the methods provided by the above-mentioned method embodiments, for example, including: s1, respectively determining normalized differential vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on time sequence remote sensing data of the target area in the first target year and the second target year; s2, respectively fitting the NDVI time sequence data of the target area in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target area in the first target year and a second NDVI time sequence curve of the target area in the second target year; s3, determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determining the annual change detection result of the target area in the first target year and the second target year based on the DTW distance value.
The logic instructions in memory 302 may be implemented in software functional units and stored in a computer readable storage medium when sold or used as a stand-alone article of manufacture. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
On the basis of the foregoing embodiments, embodiments of the present invention further provide a non-transitory computer-readable storage medium storing computer instructions that cause the computer to execute the method provided by the foregoing method embodiments, for example, including: s1, respectively determining normalized differential vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on time sequence remote sensing data of the target area in the first target year and the second target year; s2, respectively fitting the NDVI time sequence data of the target area in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target area in the first target year and a second NDVI time sequence curve of the target area in the second target year; s3, determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determining the annual change detection result of the target area in the first target year and the second target year based on the DTW distance value.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. A method for detecting change of land cover annual ring is characterized by comprising the following steps:
respectively determining normalized differential vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on the time sequence remote sensing data of the target area in the first target year and the second target year; the time sequence remote sensing data is obtained based on a Landsat satellite;
respectively fitting NDVI time sequence data of the target region in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target region in the first target year and a second NDVI time sequence curve of the target region in the second target year; the preset remote sensing time sequence model is a continuous function model for fitting discrete NDVI time sequence data;
determining a DTW distance value between the first NDVI timing curve and the second NDVI timing curve based on a Dynamic Time Warping (DTW) algorithm, and determining an annual change detection result of the target region in the first target year and the second target year based on the DTW distance value;
the preset remote sensing time sequence model is a function model determined by a sine term, a cosine term, a constant term, a primary term and a secondary term;
the specific representation form of the preset remote sensing time sequence model is as follows:
wherein, a0Is a constant term of1、a2、a3And a4Respectively cosine term coefficient and sine term coefficientThe first term coefficient and the second term coefficient are constants; x is a time parameter in days in the first or second target year, and T is the number of days in the first or second target year;
NDVI timing data corresponding to x.
2. The land cover annual variation detection method of claim 1, further comprising, prior to determining the NDVI timing data for the target area at the first target year and the second target year:
respectively preprocessing the time sequence remote sensing data of the target area in the first target year and the second target year;
the pre-treatment at least comprises: radiation calibration, atmospheric correction and geometric correction.
3. The land cover annual variation detection method according to claim 1 or 2, wherein said determining a DTW distance value between said first NDVI timing curve and said second NDVI timing curve based on a dynamic time warping, DTW, algorithm, comprises:
determining the shortest normalization path of the first NDVI time sequence curve and the second NDVI time sequence curve through the DTW algorithm according to the first NDVI time sequence curve and the second NDVI time sequence curve, and calculating the length value of the shortest normalization path;
and calculating the DTW distance value based on the length value.
4. A land cover annual variation detection method according to claim 3, characterized in that the length value of said shortest normalized path is calculated by the following formula:
D(i,j)＝Dist(i,j)+min{D(i-1,j),D(i,j-1),D(i-1,j-1)}
wherein i is the ith time coordinate in the first NDVI timing curve, and j is the second NDVIThe j-th time coordinate in the sequence curve, i is greater than or equal to 1, j is greater than or equal to 1, Dist (i, j) ═ xi-yjAnd has D (1,1) ═ Dist (1,1), xiIs the ith time coordinate value, y, in the first NDVI timing curvejIs the jth time coordinate value in the second NDVI timing curve.
5. The land cover annual variation detection method according to claim 1 or 2, wherein said determining the annual variation detection result of said target area in said first target year and said second target year based on said DTW distance value comprises:
comparing the DTW distance value to a change detection threshold;
and if the DTW distance value is judged to be larger than the change detection threshold value, determining that the annual change detection result is changed.
6. A land cover annual variation detection system, comprising:
the NDVI time sequence data acquisition module is used for respectively determining normalized difference vegetation index NDVI time sequence data of a target area in a first target year and a second target year based on time sequence remote sensing data of the target area in the first target year and the second target year; the time sequence remote sensing data is obtained based on a Landsat satellite;
the NDVI time sequence curve acquisition module is used for respectively fitting the NDVI time sequence data of the target area in the first target year and the second target year based on a preset remote sensing time sequence model, and determining a first NDVI time sequence curve of the target area in the first target year and a second NDVI time sequence curve of the target area in the second target year; the preset remote sensing time sequence model is a continuous function model for fitting discrete NDVI time sequence data;
an annual change detection result acquisition module, configured to determine, based on a Dynamic Time Warping (DTW) algorithm, a DTW distance value between the first NDVI timing curve and the second NDVI timing curve, and determine, based on the DTW distance value, an annual change detection result of the target region in the first target year and the second target year;
the preset remote sensing time sequence model is a function model determined by a sine term, a cosine term, a constant term, a primary term and a secondary term;
the specific representation form of the preset remote sensing time sequence model is as follows:
wherein, a0Is a constant term of1、a2、a3And a4The coefficient of cosine term, the coefficient of sine term, the coefficient of primary term and the coefficient of secondary term are respectively constant; x is a time parameter in days in the first or second target year, and T is the number of days in the first or second target year;
NDVI timing data corresponding to x.
7. An electronic device, comprising:
at least one processor, at least one memory, a communication interface, and a bus; wherein the content of the first and second substances,
the processor, the memory and the communication interface complete mutual communication through the bus;
the memory stores program instructions executable by the processor to invoke to perform the method of annual change in land cover according to any of claims 1 to 5.
8. A non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method of detecting annual change in land cover according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811222022.7A CN109446962B (en) | 2018-10-19 | 2018-10-19 | Land cover annual change detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811222022.7A CN109446962B (en) | 2018-10-19 | 2018-10-19 | Land cover annual change detection method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109446962A CN109446962A (en) | 2019-03-08 |
| CN109446962B true CN109446962B (en) | 2020-12-01 |
Family
ID=65547002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811222022.7A Expired - Fee Related CN109446962B (en) | 2018-10-19 | 2018-10-19 | Land cover annual change detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109446962B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111582554B (en) * | 2020-04-17 | 2024-02-23 | 北京市农林科学院信息技术研究中心 | Crop growth prediction method and system |
| CN117451631A (en) * | 2020-06-17 | 2024-01-26 | 国家基础地理信息中心 | Cultivated land change detection method, cultivated land change detection device, electronic equipment and storage medium |
| CN112507858B (en) * | 2020-12-03 | 2023-03-24 | 中国科学院东北地理与农业生态研究所 | Ecological system attribute component composition structure description method based on remote sensing vegetation index |
| CN116912698B (en) * | 2023-09-04 | 2023-12-08 | 中国地质大学（武汉） | Land coverage change monitoring method and device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512616A (en) * | 2015-11-27 | 2016-04-20 | 吉林大学 | Method for rebuilding remote sensing monitoring data set by eliminating atmospheric noise effect |
| CN106096630A (en) * | 2016-06-03 | 2016-11-09 | 福州大学 | A kind of integrated pixel and the Oryza sativa L. method for detecting automatic variation of object |
| CN106897707A (en) * | 2017-03-02 | 2017-06-27 | 苏州中科天启遥感科技有限公司 | Characteristic image time series synthetic method and device based in multi-source points |
| CN107463775A (en) * | 2017-07-24 | 2017-12-12 | 福州大学 | Vegetation based on more Indices variation tendencies is lost in whereabouts recognition methods |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102176242A (en) * | 2011-02-01 | 2011-09-07 | 环境保护部卫星环境应用中心 | Method for removing cloud noise effects in normalized difference vegetation index (NDVI) time sequence image |
| US10685408B2 (en) * | 2015-03-27 | 2020-06-16 | Omniearth, Inc. | System and method for predicting crop yield |
-
2018
- 2018-10-19 CN CN201811222022.7A patent/CN109446962B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512616A (en) * | 2015-11-27 | 2016-04-20 | 吉林大学 | Method for rebuilding remote sensing monitoring data set by eliminating atmospheric noise effect |
| CN106096630A (en) * | 2016-06-03 | 2016-11-09 | 福州大学 | A kind of integrated pixel and the Oryza sativa L. method for detecting automatic variation of object |
| CN106897707A (en) * | 2017-03-02 | 2017-06-27 | 苏州中科天启遥感科技有限公司 | Characteristic image time series synthetic method and device based in multi-source points |
| CN107463775A (en) * | 2017-07-24 | 2017-12-12 | 福州大学 | Vegetation based on more Indices variation tendencies is lost in whereabouts recognition methods |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109446962A (en) | 2019-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109446962B (en) | Land cover annual change detection method and system | |
| Bastin et al. | Aboveground biomass mapping of African forest mosaics using canopy texture analysis: toward a regional approach | |
| US8594369B2 (en) | Automatic detection of fires on earth's surface and of atmospheric phenomena such as clouds, veils, fog or the like, by means of a satellite system | |
| CN112348812B (en) | Forest stand age information measurement method and device | |
| Valero et al. | Automated location of active fire perimeters in aerial infrared imaging using unsupervised edge detectors | |
| Zhang et al. | Extraction of tree crowns damaged by Dendrolimus tabulaeformis Tsai et Liu via spectral-spatial classification using UAV-based hyperspectral images | |
| CN107610050B (en) | High-spatial-temporal-resolution land surface temperature fusion method based on multi-source satellite data | |
| S Bhagat | Use of remote sensing techniques for robust digital change detection of land: A review | |
| Belenok et al. | Investigating anthropogenically transformed landscapes with remote sensing | |
| Wang et al. | Object-oriented change detection approach for high-resolution remote sensing images based on multiscale fusion | |
| Chen et al. | Mapping double-cropped irrigated rice fields in Taiwan using time-series Satellite Pour I'Observation de la Terre data | |
| CN112765556A (en) | Method for evaluating tree age of new-born mangrove forest based on intensive time remote sensing data | |
| CN116295285A (en) | Shallow sea water depth remote sensing inversion method based on region self-adaption | |
| Azevedo et al. | Shadow detection using object area-based and morphological filtering for very high-resolution satellite imagery of urban areas | |
| Murray et al. | The novel use of proximal photogrammetry and terrestrial LiDAR to quantify the structural complexity of orchard trees | |
| Lei et al. | A novel algorithm of individual tree crowns segmentation considering three-dimensional canopy attributes using UAV oblique photos | |
| Puletti et al. | Enhancing wall-to-wall forest structure mapping through detailed co-registration of airborne and terrestrial laser scanning data in Mediterranean forests | |
| Byun et al. | Relative radiometric normalization of bitemporal very high-resolution satellite images for flood change detection | |
| CN117557897A (en) | Lodging monitoring method and device for target crops, electronic equipment and storage medium | |
| CN116910684A (en) | Method for constructing and monitoring regional drought remote sensing comprehensive index | |
| Ouerghemmi et al. | Urban vegetation mapping using hyperspectral imagery and spectral library | |
| JP6798337B2 (en) | Plant discrimination device, plant discrimination method and computer program for plant discrimination | |
| CN113657275B (en) | Automatic detection method for forest and grass fire points | |
| KR20230147893A (en) | System and method for improving detection of vegetation index | |
| CN114639014A (en) | NDVI normalization method based on high-resolution remote sensing image |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20201201 Termination date: 20211019 |