JP2011138356A - Paddy rice crop status tracking system, paddy rice crop status tracking method and paddy rice crop status tracking program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the following problem: when tracking a paddy rice crop status based on an image photographed by a synthetic aperture radar (SAR), it is difficult to secure objectivity of a threshold value for determining a flooded field. <P>SOLUTION: In this paddy rice crop status tracking system, a cluster analysis part 60 acquires an SAR image of an observation target area about two seasons or three seasons including a rice planting season wherein water is fed to a field and wherein seedings are transplanted to the field, and at least one of a field preparation season before the water feeding and a paddy rice growth season wherein paddy rice comes into a state that the paddy rice grows from the seedings, performs cluster analysis with a set of pixel values in respective points of the observation target area as a sample, and classifies the set into five classes or more. A field decision part 70 decides whether or not a field of interest is a paddy rice field based on whether or not the number of each class of the samples inside the field of interest present in the observation target area satisfies a prescribed discriminant related to magnitude relationship among the classes. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a paddy rice planting state grasping system, a paddy rice planting state grasping method, and a paddy rice planting state grasping program that use a remote sensing technique to grasp a field where paddy rice has been planted.

  Conventionally, rice production has been estimated and predicted by conducting a field survey of paddy rice planting in the field. However, it takes a lot of labor to conduct a wide-range survey of rice cultivation at the national level and prefectural level by measuring from the ground.

  From a global perspective, in some areas, methods have been developed and used to determine the rice planting status based on optical remote sensing data obtained from satellites and aircraft. However, optical remote sensing cannot be observed when clouds are present. Therefore, in Japan where rice planting, which is an important time for rice cultivation, is in the rainy season, there has been a problem that stable observation is difficult with optical remote sensing.

  In this regard, remote sensing using a synthetic aperture radar (SAR) can observe the ground surface through clouds. Therefore, in recent years, research on techniques for grasping the rice planting status using SAR images has been conducted. For example, the following non-patent document 1 shows a technique for detecting a paddy field by utilizing the difference in the backscattering intensity of SAR between two periods of the flooded period and the other period. However, since SAR has many unique phenomena compared to conventional optical sensors and is complicated to handle, paddy field observation using SAR images is still not sufficiently informed.

Naoki Ishizuka, “Estimating paddy rice planting area with high accuracy using microwave satellite imagery and geographic information system”, Agricultural Environmental Technology Research Institute Innovative Agricultural Technology Acquisition Training Text 2006, [online] [ Search on November 18, 2009] Internet <URL: http://www.niaes.affrc.go.jp/techdoc/inovlec2006/4_ishitsuka.pdf>

  In the prior art, the objectivity of the sample acquisition method for determining the threshold value for determining the flooded field and the threshold value determination method remains as a problem.

  The present invention has been made to solve the above problems, and an object thereof is to provide a paddy rice planting situation grasping system, a rice planting situation grasping method, and a rice planting situation grasping program that do not depend on the threshold method.

  The paddy rice planting state grasping system according to the present invention is provided in an observation target region based on a radar image which is photographed by a synthetic aperture radar using an X-band microwave and has a pixel value corresponding to the intensity of a backscattered wave. A paddy rice planting state grasping system for grasping a paddy rice field where paddy rice has been planted, wherein the rice planting period from entering the paddy field to planting seedlings and preparing the field before entering the paddy rice field The radar image of the observation area taken at a plurality of periods including at least one of the growing season and at least one of the long-term paddy rice growth in which the rice in the paddy field is in a state where leaves are expanded more than the seedling at the time of planting A sample having a plurality of sample points set in the observation target region and having coordinates defined by the set of pixel values at the sample points of each analysis target image Cluster analysis means for classifying the sample into a plurality of classes by cluster analysis, and whether the number of the samples corresponding to the sample points in the field of interest satisfies the predetermined discriminant regarding the magnitude relationship between the classes And a field determination means for determining whether or not the field of interest is the paddy rice field.

  The paddy rice planting status grasping system further includes a discriminant determining unit that obtains the discriminant based on the number of the samples for each class with respect to a reference field that is determined to be the paddy rice field in the observation target region. It can have.

  A preferred aspect of the present invention is a paddy rice planting state grasping system in which the cluster analysis is an ISODATA method in which the number of initial classes is set to 5 or more.

  The paddy rice planting state grasping method according to the present invention is observed using a computing device based on a radar image having a pixel value corresponding to the intensity of a backscattered wave, which is photographed by a synthetic aperture radar using an X-band microwave. A paddy rice planting status grasping method for grasping a paddy rice field where paddy rice has been planted among fields provided in a target area, the rice planting period from entering the paddy field to planting seedlings, and entering the paddy rice field The observation target region taken at a plurality of times including a previous field preparation period and at least one of paddy rice growth periods in which leaves are expanded from the seedlings at the time of planting in the paddy rice field A radar image is set as an analysis target image, a plurality of sample points are set in the observation target region, and the coordinates defined by the set of pixel values at the sample points of each analysis target image are provided. A cluster analysis step for classifying samples to be classified into a plurality of classes by cluster analysis, and the number of the samples corresponding to the sample points in the field of interest for each class satisfies a predetermined discriminant regarding the magnitude relationship between the classes And a field determination step for determining whether or not the field of interest is the paddy rice field.

  The paddy rice planting situation grasping program according to the present invention is based on a radar image captured by a synthetic aperture radar using X-band microwaves and having a pixel value corresponding to the intensity of a backscattered wave. A paddy rice planting status grasping program for functioning as a system for grasping a paddy rice field planted with rice among the fields provided in the paddy field, the rice planting period from entering the paddy rice field until planting seedlings And a plurality of periods including a field preparation period before entering the paddy rice field and at least one of the paddy rice growing period in which the paddy rice in the paddy field is expanded from the seedling at the time of planting. The radar image of the observed region to be analyzed is set as an analysis target image, and a plurality of sample points are set in the observation target region. A cluster analysis function for classifying a sample having coordinates defined by the set of pixel values at the sample point of an elephant image into a plurality of classes by cluster analysis, and the sample corresponding to the sample point in the field of interest A field determination function for determining whether or not the field of interest is the paddy rice field is realized based on whether or not the number of each class satisfies a predetermined discriminant regarding the magnitude relationship between the classes.

  According to the present invention, it is possible to grasp the rice cultivation state more objectively without using the threshold method. In addition, since it is possible to observe the ground without being affected by clouds, it is possible to grasp the planting situation in the rainy season from rice planting to the growing period in Japan. In addition, appropriate cultivation management and production management can be determined. In addition, it is possible to grasp a wide range of fields in a lump with little effort.

It is a block diagram which shows the structure of the outline of the paddy rice planting condition grasping | ascertainment system which is embodiment of this invention. It is a graph which shows the measurement result of the backscattering coefficient by HH mode in a rice field, a soybean field, and a reservoir. It is a graph which shows the measurement result of the backscattering coefficient by VV mode in a rice field, a soybean field, and a reservoir. It is a functional block diagram of the paddy rice cultivation situation grasping system which is an embodiment of the present invention. It is a graph showing the feature-value of each of a rice field, a soybean field, and a reservoir at the time of setting a class number to 5 and analyzing a measurement result by HH mode. It is a graph showing the feature-value of each of a rice field, a soybean field, and a reservoir at the time of setting the number of classes to 5 and analyzing the measurement result by VV mode.

  Hereinafter, embodiments of the present invention (hereinafter referred to as embodiments) will be described with reference to the drawings.

  FIG. 1 is a block diagram illustrating a schematic configuration of a paddy rice planting state grasping system 2 according to the embodiment. This system classifies whether or not a field provided in the observation target region is a paddy rice field planted with rice based on a radar image (SAR image) of the observation target region captured by the SAR. The system includes an arithmetic processing device 4, a storage device 6, an input device 8, and an output device 10. As the arithmetic processing unit 4, it is possible to make dedicated hardware for performing the processing of this system. However, in this embodiment, the arithmetic processing unit 4 is constructed using a computer and a program executed on the computer. Is done.

  The arithmetic processing unit 4 is composed of a CPU (Central Processing Unit) of a computer, and functions as, for example, the cluster analysis unit 12, the field determination unit 14, and the discriminant determination unit 16, and further performs various functions described later with reference to FIG. Realize the part.

  The storage device 6 stores a program for causing the arithmetic processing unit 4 to function as each of the means 12 to 16 and the like, other programs, and various data necessary for processing of the present system. For example, the storage device 6 is used to hold a SAR image to be analyzed and a cluster image obtained by analysis.

  The input device 8 is a keyboard, a mouse, or the like, and is used for a user to operate the system.

  The output device 10 is a display, a printer, or the like, and is used for displaying the analysis result of the rice planting situation obtained by the present system to the user by screen display, printing, or the like. In addition, data may be output to a device other than the present system.

The SAR image used in the present system is obtained by irradiating the ground with X-band microwaves and observing the backscattered wave, and has a pixel value corresponding to the intensity of the backscattered wave. For example, a backscattering coefficient σ ⁰ expressed by the following equation can be used as a pixel value.
σ ⁰ [dB] = ₁₀ log ₁₀ (k · | DN | ² · sinθ _loc )

Here, σ ⁰ is expressed in decibel values, k is a coefficient for calibration and processor scaling, DN is an amplitude of backscattering, and θ _loc is an incident angle.

  In the present embodiment, an SAR image based on imaging data in an X band (wavelength: 3.1 cm) by TerraSAR-X, which is a SAR satellite, is used. The imaging data used was acquired in the HH mode in which both transmission and reception are performed with horizontal polarization, and acquired in the VV mode in which both transmission and reception are performed with vertical polarization. When the TerraSAR-X shooting mode is a high resolution spotlight, data obtained by shooting a 10 km × 5 km area with a resolution of 2.2 m is acquired. TerraSAR-X can shoot the same area in an 11-day cycle.

  FIG. 2 is a graph showing the measurement results of the backscattering coefficient in a rice field, soybean field, and reservoir in a certain year in the HH mode. Moreover, FIG. 3 is a graph which shows the measurement result by the VV mode of the backscattering coefficient in a paddy rice field, a soybean field, and a reservoir. 2 and 3, the vertical axis represents the backscattering coefficient, the horizontal axis represents the photographing date, and the change over time before and after the rice planting period is represented. As for the paddy rice field, three cases (practice, early planting, late planting) in which the dry rice transplanting times are different are shown, and change curves 30 (denoted by Δ), 32 (denoted by □), 34 in FIG. (Represented by ○) represents the change in the backscattering coefficient in the case of conventional, early planting, and late planting, respectively, and the change curves 31 (represented by Δ), 33 (represented by □), and 35 (represented by ○) (Notation) represents the change in the backscattering coefficient in the case of customary, early planting, and late planting, respectively. Also, the change curve 36 in FIG. 2 and the change curve 37 in FIG. 3 (both are indicated by “x”) are soybean fields, the change curve 38 in FIG. 2 and the change curve 39 in FIG. 3 (both are indicated by “+”) are the backscattering coefficients of the reservoir. Represents changes.

  In the seven shootings of the HH mode shown in FIG. 2 and the VV mode shown in FIG. 3, the backscattering coefficient of the soybean field changes within the range of −10 to −15 dB, and the backscattering coefficient of the reservoir is around −20. Has changed. In the early-planted paddy field, the water was introduced in late April and transplanted in early May. In the conventional paddy field, the water entered in early May and transplanted in the middle. In the late-planted paddy field, the water entered and transplanted in late May. Was done. Each of these three paddy rice fields has a backscattering coefficient similar to that of the soybean field before entering the water, and the backscattering coefficient decreases corresponding to the time of transplantation from the time of entering water and approaches the level of the backscattering coefficient of the reservoir. After that, the backscattering coefficient increased with time and again reached the same level as in the soybean field. The characteristics of this change were common to the HH mode and VV mode. In paddy rice fields, it is considered that the unevenness of the ground surface affects the backscattering and the backscattering coefficient becomes large during the rice paddying stage before water entry. In addition, the ground surface is covered with water during the period from water entry to transplantation. It is considered that specular reflection occurs due to this flooded state, and the backscattering coefficient becomes small. Immediately after transplanting, the rice is in the state of seedlings with few leaves, and the water surface appears relatively well between them, but as the rice grows and the number of leaves increases, the rice affects the backscattering coefficient. Given this, the backscattering coefficient is thought to increase.

  FIG. 4 is a functional block diagram of this system. The interface (I / F) unit 50 has a function of taking a captured SAR image into the present system. The user sets the shooting time of the SAR image based on the paddy rice cultivation and the shooting cycle of the satellite, and the SAR image of the observation target area at the set shooting times is captured by the satellite. The present system captures the SAR image thus taken.

  The captured time-series SAR images are stored in the SAR image holding unit 52. For example, the storage device 6 is used as a SAR image holding unit. The alignment processing unit 54 is based on map data 56 such as topographic maps and reference points (road intersections, buildings, etc.) with clear latitude and longitude so that SAR images of a plurality of periods can be superimposed on the map. A geometric (position) correction process is performed on the SAR image.

  The SAR image noise processing unit 58 processes noise of the SAR image. For example, the SAR image noise processing unit 58 performs a filter process on the SAR image after the alignment process, and removes noise such as speckle noise from the SAR image. For example, a median filter, a Frost filter, a Lee filter, or the like is used as the filter processing, and these are selected according to the noise characteristics of the SAR image. Note that the processing order of the alignment processing unit 54 and the SAR image noise processing unit 58 may be reversed.

  The cluster analysis unit 60 performs cluster analysis on the SAR image that has been subjected to noise removal processing. The SAR image that is the subject of cluster analysis is taken at one or both of those taken during the rice planting period, the field preparation period that is the period before the rice planting period, and the rice growing period that is the period after the rice planting period. Are a plurality of images. From the viewpoint of reducing time and cost, it is preferable to use less data. From this viewpoint, the image to be analyzed is taken in one of the rice planting period and either the field preparation period or the rice cultivation period. It is possible to make a total of 3 images, that is, a total of 2 images with 1 image that has been made, or 1 image of the rice planting period and 2 images that have been photographed during the field preparation period and the rice growing period. However, images in the rice planting period, the field preparation period, and the paddy rice growth period may be added to the two images of the two periods or the three images of the three periods, and these may be used as analysis target images.

  Here, the rice planting period corresponds to a period from entering the paddy rice field to planting rice seedlings. The field preparation period is a period before entering the paddy rice field, such as a rice paddying period. The long-term paddy rice growth is more than the seedling state at the time of planting, and the transplanted rice is rooted and the number of leaves (leaf age) is increased from the time of transplanting. Refers to the time when Since the number of leaves increases by one in about one week, the long-term paddy rice growth can be made a period after about one week has passed since rice planting. More preferably, an SAR image at a time when the growth situation of rice is greatly changed from immediately after rice planting is used as an image of the paddy rice growth period. From this point of view, in the present embodiment, two or three cycles after the TerraSAR-X shooting period is selected as the shooting period of the paddy rice growth period from the time of shooting at the rice planting period. Specifically, the SAR image of April 20 is used as the image in the field preparation period, the SAR image on May 23 is used as the image in the rice planting period, and the SAR image on June 25 is used as the image in the paddy rice growth period. Using.

The cluster analysis unit 60 performs cluster analysis on samples defined at a plurality of sample points set in the observation target region. A sample is defined by a set of pixel values of each analysis target image at a corresponding sample point. For example, when the analysis target images are two images I ₁ and I ₂ obtained in two periods (one of the rice planting period and the field preparation period or the rice growing period), the sample S at the sample point P is , And is defined by a set (d ₁ , d ₂ ) of pixel values d ₁ , d ₂ at the sample points P of the images I ₁ , I ₂ . In addition, when the analysis target images are three images I ₁ , I ₂ , and I ₃ obtained in three periods (rice planting period, field preparation period, and paddy rice growth period), the sample S includes images I ₁ , It is defined by a set (d ₁ , d ₂ , d ₃ ) of pixel values d ₁ , d ₂ , d ₃ at each sample point P of I ₂ , I ₃ . Note that the alignment processing unit 54 associates the sample points P defined in the observation target region with the pixels of the images I ₁ , I ₂ , and I ₃ . In the present embodiment, each pixel of the SAR image is associated with a sample point.

  The cluster analysis unit 60 uses the ISODATA method as a cluster analysis method. Similarly to the ISODATA method, the k-means method which is a non-hierarchical method and an unsupervised classification method is also suitable. A supervised classification method may be used. A non-hierarchical method is suitable for cluster analysis in an image where the number of samples can be enormous, from the viewpoint of reducing the amount of calculation. However, if the processing power of the arithmetic processing unit 4 is high, a hierarchical method can be used. is there.

When the ISODATA method is used, the initial class number is set to 5 or more. The center of the initial class is set based on the conventional method. For example, when the analysis target image is three images, the average μ _j and the standard deviation σ _j of the pixel values of each analysis target image I _j (j = 1, 2, 3) are obtained, and the center or centroid (α ₁ , alpha _2, equally spaced coordinate values alpha _j of alpha ₃₎ in the range of _{μ j -σ j ~μ j + σ} j.

  Various backscattering components contribute to the backscattering intensity represented by each pixel value of the SAR image. For example, a specular scattering component due to the water surface, a scattering component due to the surface roughness of rice or soil, a corner reflection component of two surfaces due to buildings, and the like. Depending on the strength of such components, samples are classified into multiple classes by cluster analysis. The observation area includes not only paddy rice fields but also various types of fields such as soybean and other crop fields, reservoirs, etc., and the cluster analysis classification results also include the contribution of backscattering from them. The influence of ingredients appears.

  Here, class identification numbers k are assigned in ascending order from the smallest average value of the distance from the center of the class to the sample belonging to the class. For example, the class having the smallest average distance is “class 1”.

  A cluster image having a pixel value corresponding to the class at each sample point is stored in the cluster image holding unit 62, and is output to a display device or the like when, for example, an analysis result of the rice planting situation is displayed.

  The field frame determination unit 64 has a function of determining the position of a field that is a planting unit in the observation target region, and obtains a field frame that represents the range of each field. For example, the field frame determination unit 64 determines the field frame based on an aerial photograph or a numerical map issued by the Geographical Survey Institute. Note that the field frame data may not be extracted by the present system, but the field frame data extracted in advance may be input from the outside using a GIS (Geographic Information System) or the like.

The farm field feature amount calculation unit 66 superimposes the farm field frame on the cluster image, and extracts the cluster image in the farm field frame for each farm field. And the ratio of the number of samples of each class in the extracted cluster image of a field is calculated | required as a feature-value. Specifically, in the present embodiment in which pixels corresponding to the sample, wherein the amount of the fraction (C k _{/ N)} of the pixel number C _k of class k to the total number of pixels N constituting the extracted field clusters image chi ( k).

  The paddy rice field and other fields have different scattering component changes in a plurality of shooting periods in addition to the scattering component difference in the SAR image at the same shooting time. In particular, as described above, in the paddy rice field, a SAR image having a large influence of reflection from the flooded surface is obtained during the rice planting period, and a SAR image having no or little influence from the flooded surface is obtained during the field preparation period and the paddy rice growing period. In this respect, there is a difference from other fields in the change of the scattering component in a plurality of imaging periods. For these reasons, there may be a difference in the magnitude relationship between the feature quantity classes between the paddy rice field and other fields.

  FIG. 5 is a graph showing the characteristic quantities of the rice field, soybean field, and reservoir when the number of classes is set to 5 and the measurement results in the HH mode are analyzed. FIG. 6 is a graph showing the characteristic amounts of the rice field, soybean field, and reservoir when the number of classes is set to 5 and the measurement result in the VV mode is analyzed. 5 (a) and 6 (a) are the conventional paddy rice fields, FIG. 5 (b) and FIG. 6 (b) are the soybean fields, and FIG. 5 (c) and FIG. 6 (c) are the features of the reservoir. The feature quantity χ (k) is expressed as a percentage on the vertical axis, and classes are arranged on the horizontal axis. As shown in FIGS. 5A and 6A, in the paddy rice field, χ (1) and χ (3) are relatively large and χ (2) is small. As shown in FIGS. 5B and 6B, in the soybean field, χ (2) is larger than other feature values. As shown in FIGS. 5C and 6C, in the reservoir, χ (1) is larger than other feature values.

  The discriminant determination unit 68 determines a discriminant for determining whether or not the field of interest is a paddy rice field with respect to the magnitude relationship between the classes of these feature values. Specifically, in this system, a reference field that is known in advance to be a paddy rice field is designated by a user, a feature amount of a cluster image of the designated field is calculated, and a discriminant formula is calculated based on the feature amount. Ask for. The discriminant is not determined by the present system, and a discriminant determined in advance by a similar method may be set from the outside.

For example, with respect to the feature amounts shown in FIGS. 5 and 6, the discriminant equation indicating the paddy rice field is the following equation.
χ (1)> χ (2)
χ (2) <χ (3) (1)
χ (3)> χ (4)

A similar relational expression for feature quantities is
χ (1) <χ (2)
χ (2)> χ (3) (2)
χ (3)> χ (4)
In the reservoir,
χ (1)> χ (2)
χ (2)> χ (3) (3)
χ (3)> χ (4)
Thus, the equations (2) and (3) are different from the equation (1).

  The farm field determination unit 70 checks whether each farm field in the observation target region satisfies the discriminant, and determines whether the farm field is a paddy rice field. The determined results are collected by the output unit 72 into a classification diagram of the paddy rice field, or output to a list of planted items in each field. Further, a classification chart of paddy rice fields, a list of planted items, and the like can be presented to the user by the output device 10.

  For example, the field determination unit 70 determines that a field that satisfies the above-described discriminant (equation (1)) is a paddy field, and determines that a field that does not satisfy is not a paddy field.

  According to this system, it becomes possible to grasp the planting situation at the same time as the rice planting period by the analysis based on the two images of the field preparation period and the rice planting period. In addition, it is possible to grasp the planting situation at a certain time with the images of the two periods, and the cost of the field survey can be reduced. Further, since the number of images is small, it is possible to reduce cost cost and time cost. Furthermore, since the planting status is grasped by a certain work, objectivity can be maintained and human errors can be reduced.

  2 Paddy rice planting status grasping system, 4 arithmetic processing device, 6 storage device, 8 input device, 10 output device, 12 cluster analysis means, 14 field determination means, 16 discriminant determination means, 50 interface part, 52 SAR image holding part, 54 registration processing unit, 56 map data, 58 SAR image noise processing unit, 60 cluster analysis unit, 62 cluster image holding unit, 64 field frame determination unit, 66 field feature amount calculation unit, 68 discriminant determination unit, 70 field determination Part, 72 output part.

Claims (5)

A paddy rice field in which paddy rice is planted among the fields provided in the observation area based on a radar image captured by a synthetic aperture radar using X-band microwaves and having a pixel value corresponding to the intensity of the backscattered wave Rice paddy cultivation system
The rice planting period from entering the paddy rice field to planting seedlings, the field preparation period before entering the paddy rice field, and the paddy rice in the paddy rice field having leaves spread more than the seedlings when planting The radar image of the observation target region taken at a plurality of periods including at least one of the rice growing period is set as an analysis target image, and a plurality of sample points are set in the observation target region, and each analysis target image Cluster analysis means for classifying samples having coordinates defined by the set of pixel values at the sample points into a plurality of classes by cluster analysis;
Whether the field of interest is the rice field based on whether the number of the samples corresponding to the sample points in the field of interest satisfies a predetermined discriminant regarding the magnitude relationship between the classes A field determination means for determining whether or not,
A rice planting situation grasping system characterized by having. In the paddy rice planting status grasping system according to claim 1,
Characterized in that it has discriminant determining means for determining the discriminant based on the number of the samples for each class with respect to a reference field that is determined to be the paddy rice field in the observation target region. Situation grasp system. In the paddy rice planting status grasping system according to claim 1 or claim 2,
The cluster analysis is an ISODATA method in which the initial number of classes is set to 5 or more. Based on a radar image having a pixel value corresponding to the intensity of the backscattered wave, which is taken by a synthetic aperture radar using X-band microwaves, a paddy rice is selected from among the fields provided in the observation target area using an arithmetic unit. A method for grasping the state of rice cultivation to grasp the planted rice field,
The rice planting period from entering the paddy rice field to planting seedlings, the field preparation period before entering the paddy rice field, and the paddy rice in the paddy rice field having leaves spread more than the seedlings when planting The radar image of the observation target region taken at a plurality of periods including at least one of the rice growing period is set as an analysis target image, and a plurality of sample points are set in the observation target region, and each analysis target image A cluster analysis step of classifying samples having coordinates defined by the set of pixel values at the sample points into a plurality of classes by cluster analysis;
Whether the field of interest is the rice field based on whether the number of the samples corresponding to the sample points in the field of interest satisfies a predetermined discriminant regarding the magnitude relationship between the classes A field determination step for determining whether or not,
A method for grasping the rice planting status, characterized by comprising: The paddy rice is planted in the field provided in the observation area based on the radar image that is captured by a synthetic aperture radar using X-band microwaves and has a pixel value corresponding to the intensity of the backscattered wave. A paddy rice cultivation situation grasping program for functioning as a system for grasping paddy rice fields,
On that computer,
The rice planting period from entering the paddy rice field to planting seedlings, the field preparation period before entering the paddy rice field, and the paddy rice in the paddy rice field having leaves spread more than the seedlings when planting The radar image of the observation target region taken at a plurality of periods including at least one of the rice growing period is set as an analysis target image, and a plurality of sample points are set in the observation target region, and each analysis target image A cluster analysis function for classifying samples having coordinates defined by the set of pixel values at the sample points into a plurality of classes by cluster analysis;
Whether the field of interest is the rice field based on whether the number of the samples corresponding to the sample points in the field of interest satisfies a predetermined discriminant regarding the magnitude relationship between the classes A field determination function for determining whether or not,
Rice paddy cultivation situation grasping program characterized by realizing.

Images