JP6413445B2 - Plant discrimination device, plant discrimination method, and plant discrimination program - Google Patents

Description

  The present invention relates to a plant determination device and the like.

  As a method for investigating the type of plant, there are a first investigation method and a second investigation method. The first investigation method is a method in which an identifier performs a site survey and visually discriminates the situation of the site. The second investigation method is a method of identifying using a photograph or an image taken from an artificial satellite or an aircraft. This second investigation method is called remote sensing. The first investigation method and the second investigation method are used individually or in combination.

  In recent years, a satellite equipped with a hyperspectral sensor capable of measuring a band more than 10 times the conventional multispectrum has been launched as a global environment satellite, and remote sensing is performed using the measurement result of the hyperspectral sensor.

  The spectral data obtained by the hyperspectral sensor is data having wavelength information and light intensity information for each pixel of the image. Since the reflection spectrum is unique to each tree species, the tree species can be identified using hyperspectral data. For example, the reference spectrum of the target tree is compared with the spectrum data to be determined to determine the similarity for each pixel, and a pixel having a high similarity is determined as a region where the target tree is located.

JP 2011-69757 A JP 2005-199154 A JP 2010-237865 A

  However, the above-described conventional technique has a problem that the accuracy of identifying a plant is low.

  When spectrum data is acquired using an artificial satellite, an aircraft, or the like, the spectrum data includes not only information on forests and farmland but also information on artifacts in urban areas. Since the characteristic of the reflection spectrum of the artificial object is similar to the characteristic of the reflection spectrum of the plant, the artificial object may be mistakenly identified as a plant, and plant discrimination accuracy is reduced.

  In one aspect, an object is to provide a plant discrimination device, a plant discrimination method, and a plant discrimination program that can increase the discrimination accuracy of a plant.

  In the first plan, the plant discrimination device has an exclusion unit. The exclusion unit uses spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region for each pixel of the image. The excluding unit multiplies the first median value by a predetermined coefficient for the set of the first median value of the light intensity in the visible light region and the second median value of the light intensity in the infrared light region corresponding to a certain pixel on the image. When the obtained value is smaller than the second median value, a certain pixel is excluded from the plant discrimination targets.

  According to one embodiment of the present invention, there is an effect that the discrimination accuracy of plants can be increased.

FIG. 1 is a functional block diagram illustrating the configuration of the plant discrimination device according to the present embodiment. FIG. 2 is a diagram illustrating an example of hyperspectral data. FIG. 3 is a diagram illustrating an example of corrected hyperspectral data. FIG. 4 is a diagram illustrating a relationship between the first median value and the second median value of each pixel. FIG. 5 is a flowchart illustrating the processing procedure of the plant discrimination device according to the present embodiment. FIG. 6 is a diagram illustrating an example of a computer that executes a plant discrimination program.

  Hereinafter, embodiments of a plant discrimination device, a plant discrimination method, and a plant discrimination program disclosed in the present application will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

  A configuration of the plant discrimination device according to the present embodiment will be described. FIG. 1 is a functional block diagram illustrating the configuration of the plant discrimination device according to the present embodiment. As illustrated in FIG. 1, the plant discrimination device 100 includes a communication unit 110, an input unit 120, a display unit 130, a storage unit 140, and a control unit 150.

  The communication unit 110 is a processing unit that performs data communication with other devices via a network. For example, the communication unit 110 corresponds to a communication device or the like. For example, the communication unit 110 outputs the received hyperspectral data 141 to the control unit 150 when the hyperspectral data 141 described later is received from another device.

  The input unit 120 is an input device that inputs various types of information to the plant discrimination device 100. For example, the input unit 120 corresponds to a keyboard, a mouse, a touch panel, or the like.

  The display unit 130 is a display device that displays various data output from the control unit 150. For example, the display unit 130 corresponds to a liquid crystal display, a touch panel, or the like.

  The interface unit 135 is an interface that is connected to a portable storage medium or the like and exchanges data with the portable storage medium. The portable storage medium corresponds to, for example, a USB (Universal Serial Bus) memory, an SD memory card, a memory stick, a compact flash (registered trademark), and the like. For example, the interface unit 135 acquires the hyperspectral data 141 stored in the portable storage medium, and outputs the acquired hyperspectral data 141 to the control unit 150.

  The storage unit 140 has hyperspectral data 141 and corrected hyperspectral data 142. The storage unit 140 corresponds to a storage device such as a semiconductor memory element such as a random access memory (RAM), a read only memory (ROM), and a flash memory.

  The hyperspectral data 141 is data including wavelength information and light intensity information for each pixel of the image. FIG. 2 is a diagram illustrating an example of hyperspectral data. Although the description is omitted in FIG. 2, the hyperspectral data 141 is configured by overlapping images for each wavelength, and the pixel, the wavelength, and the light intensity are associated with each other. For example, the wavelength range of the hyperspectral data 141 is 400 nm to 1400 nm.

  The corrected hyperspectral data 142 is hyperspectral data obtained by deleting the artifact region from the hyperspectral data 141. FIG. 3 is a diagram illustrating an example of corrected hyperspectral data. In the example shown in FIG. 3, each area 10a is determined as an artifact and is deleted. The process of deleting the artifact region from the hyperspectral data 141 will be described later.

  Returning to the description of FIG. The control unit 150 includes an acquisition unit 151, an exclusion unit 152, and a collation unit 153. The control unit 150 corresponds to an integrated device such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA). The control unit 150 corresponds to an electronic circuit such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit).

  The acquisition unit 151 is a processing unit that acquires the hyperspectral data 141 from the communication unit 110 or the interface unit 135 and registers the acquired hyperspectral data 141 in the storage unit 140.

  Based on the hyperspectral data 141, the excluding unit 152 identifies an artificial region on the image, and excludes the identified artificial region from the hyperspectral data 141, thereby generating corrected hyperspectral data 142. It is a processing unit. The excluding unit 152 registers the generated corrected hyperspectral data 142 in the storage unit 140.

  The exclusion unit 152 compares the light intensities included in the visible light region corresponding to the pixels on the image to identify the median value. The wavelength range of the visible light region is 495 nm to 570 nm. In the following description, the median value of each light intensity included in the visible light region is referred to as a first median value.

  The exclusion unit 152 compares the light intensities in the infrared light region corresponding to the pixels on the image, and specifies the median value. The wavelength range of the infrared light region is 750 nm to 1400 nm. In the following description, the median value of each light intensity included in the infrared light region is referred to as a second median value.

  The excluding unit 152 determines whether or not a value obtained by multiplying the first median by a predetermined coefficient for a set of the first median and the second median for a pixel on the image is smaller than the second median. To do. When the value obtained by multiplying the first median by a predetermined coefficient is smaller than the second median, the excluding unit 152 determines that the corresponding pixel corresponds to the artifact. The excluding unit 152 specifies the region corresponding to the artifact on the image by executing the above process on all the pixels of the image of the hyperspectral data 141.

  FIG. 4 is a diagram illustrating a relationship between the first median value and the second median value of each pixel. The horizontal axis in FIG. 4 is an axis corresponding to the first median value in the visible light region. The vertical axis is an axis corresponding to the second median value in the infrared light region. Each pixel shown in FIG. 4 is a pixel corresponding to each pixel of the hyperspectral data 141.

  A line segment 20 shown in FIG. 4 is a line segment corresponding to the equation (1). In Expression (1), x is a variable corresponding to the first median value. y is a variable corresponding to the second median value. a corresponds to a predetermined coefficient, and in the example shown in FIG. 4, the coefficient is “1.4”.

  y = ax (1)

  In FIG. 4, each pixel in which the relationship between the first median and the second median is located below the line segment 20 is a pixel corresponding to the artifact. On the other hand, each pixel in which the relationship between the first median and the second median is located above the line segment 20 is a pixel corresponding to other than the artifact.

  As illustrated in FIG. 4, the exclusion unit 152 identifies each pixel in which the relationship between the first median and the second median is located below the line segment 20 from the hyperspectral data 141, and identifies the identified pixel. Is generated, the corrected hyperspectral data 142 is generated.

  Returning to the description of FIG. The collation unit 153 is a processing unit that collates the corrected hyperspectral data 142 with a reference spectrum for each tree type to determine a tree type existing on the corrected hyperspectral data 142. The collation unit 153 outputs the determination result to the display unit 130.

  Next, a processing procedure of the plant discrimination device 100 according to the present embodiment will be described. FIG. 5 is a flowchart illustrating the processing procedure of the plant discrimination device according to the present embodiment. As shown in FIG. 5, the plant discrimination device 100 determines whether or not to start the process (step S101). When the process is not started (No at Step S101), the plant identification device 100 proceeds to Step S101 again.

  When starting the process (Yes in step S101), the excluding unit 152 of the plant discrimination device 100 specifies the first median value of the light intensity in the visible light region for each pixel (step S102). The excluding unit 152 specifies the second median value of the light intensity in the infrared light region for each pixel (step S103).

  The excluding unit 152 sets a coefficient (step S104). The excluding unit 152 identifies a pixel in which the relationship between the first median and the second median satisfies a predetermined condition (Step S105). For example, in step S <b> 105, the excluding unit 152 specifies a pixel whose value obtained by multiplying the first median value by a coefficient is smaller than the second median value.

  The exclusion unit deletes the identified pixel region from the hyperspectral data 141 and generates corrected hyperspectral data 142 (step S106). The collation unit 153 performs collation based on the corrected hyperspectral data 142 (step S107).

  Next, the effect of the plant discrimination device 100 according to the present embodiment will be described. The plant discriminating apparatus 100 has a value obtained by multiplying the first median by a predetermined coefficient for the set of the first median and the second median for the pixel on the screen of the hyperspectral data 141, than the second median. It is determined whether or not it is small. The plant discrimination device 100 excludes the corresponding pixel from the hyperspectral data 141 when the value obtained by multiplying the first median by a predetermined coefficient is smaller than the second median. A pixel that satisfies the above condition corresponds to an artificial object, and by using the hyperspectral data 141 from which the artificial object has been removed, plant discrimination accuracy can be improved.

  In the said Example, the plant determination apparatus 100 is good also as a value other than "1.4" for the value of the coefficient a. For example, the value of the coefficient a may be “1.2 to 1.6”, preferably “1.3 to 1.4”.

  Here, when the value of the coefficient a in the formula (1) is “1.4”, “1.6”, “1.2”, “1.8”, “1.0”, the hyperspectral data The results of an actual survey conducted by the inventor on the area excluded from the above will be described.

  The deletion area and the reconnaissance result when the value of the coefficient a is “1.4” will be described. As a result of examining the 30 areas actually deleted, the inventor confirmed that all of the areas were artifacts. Further, as a result of examining the area close to the deleted area, the inventor was able to confirm that all were plants.

  The deletion area and the reconnaissance result when the value of the coefficient a is “1.6” will be described. As a result of inspecting 35 areas actually deleted, the inventor confirmed that 30 places were artifacts and 5 places were plants. Further, as a result of examining the area close to the deleted area, the inventor was able to confirm that all were plants.

  The deletion area and the reconnaissance result when the value of the coefficient a is “1.2” will be described. As a result of probing 30 areas actually deleted, the inventor confirmed that 30 areas were artifacts. Note that all the deleted areas were artifacts, but it was confirmed that seven of the areas adjacent to the deleted areas were artifacts.

  The deletion area and the reconnaissance result when the value of the coefficient a is “1.8” will be described. As a result of inspecting 47 areas that were actually deleted, the inventor confirmed that 30 places were artifacts and 17 places were plants. Further, as a result of examining the area close to the deleted area, the inventor was able to confirm that all were plants.

  The deletion area and the reconnaissance result when the value of the coefficient a is “1.0” will be described. As a result of examining the 30 areas actually deleted, the inventor confirmed that 30 areas were artifacts, but 12 areas adjacent to the deleted areas were artifacts.

  Therefore, by setting the value of the coefficient a to “1.2 to 1.6”, preferably “1.3 to 1.4”, it is possible to accurately exclude the artifact from the hyperspectral data 141. It could be confirmed.

  The plant identification device 100 uses the first median value specified from the visible light region (495 nm to 570 nm) and the second median value specified from the infrared light region (750 nm to 1400 nm) to accurately identify the artifact. can do.

  In addition, although the exclusion part 152 which concerns on a present Example calculated | required the 2nd median value from the light intensity contained in 750 nm-1400 nm, it is not limited to this. For example, the second median value may be specified from the light intensity included in 450 nm to 1400 nm + α nm. The value of α is a value set by the user as appropriate.

  Moreover, the plant discrimination device 100 can obtain information for specifying the first and second median values by setting the wavelength range of the hyperspectral data 141 to 400 nm to 1400 nm.

  By the way, the plant discrimination device 100 determines that a pixel whose value obtained by multiplying the first median by a predetermined coefficient for the first and second median sets of pixels is smaller than the second median is an artifact. It was. However, it goes without saying that the plant discrimination determination unit 100 may determine, as an artifact, a pixel having a smaller first median value than a value obtained by multiplying the second median value by the inverse of a predetermined coefficient.

  Next, an example of a computer that executes a plant discrimination program that realizes the same function as the plant discrimination apparatus 100 shown in the above embodiment will be described. FIG. 6 is a diagram illustrating an example of a computer that executes a plant discrimination program.

  As illustrated in FIG. 6, the computer 200 includes a CPU 201 that executes various arithmetic processes, an input device 202 that receives data input from a user, and a display 203. The computer 200 also includes a reading device 204 that reads a program and the like from a storage medium, and an interface device 205 that exchanges data with other computers via a network. The computer 200 also includes a RAM 206 that temporarily stores various information and a hard disk device 207. The devices 201 to 207 are connected to the bus 208.

  The hard disk device 207 has an exclusion program 207a. The CPU 201 reads out the exclusion program 207a and develops it in the RAM 206. The exclusion program 207a functions as an exclusion process 206a. The processing executed by the exclusion process 206a corresponds to the processing executed by the exclusion unit 152.

  Note that the exclusion program 207a is not necessarily stored in the hard disk device 207 from the beginning. For example, each program is stored in a “portable physical medium” such as a flexible disk (FD), a CD-ROM, a DVD disk, a magneto-optical disk, and an IC card inserted into the computer 200. Then, the computer 200 may read and execute the exclusion program 207a.

  The following supplementary notes are further disclosed with respect to the embodiments including the above examples.

(Supplementary note 1) The visible light corresponding to a certain pixel on the image based on spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region for each pixel of the image For a set of a first median value of light intensity in the light region and a second median value of light intensity in the infrared light region, a value obtained by multiplying the first median value by a predetermined coefficient is greater than the second median value. A plant discrimination device comprising: an excluding unit that excludes a certain pixel from a plant discrimination target when the pixel is small.

(Additional remark 2) The said predetermined coefficient is a value from 1.2 to 1.6 or a value from 1.3 to 1.4, The plant discrimination apparatus of Additional remark 1 characterized by the above-mentioned.

(Additional remark 3) The wavelength range of the said visible region is 495 nm to 570 nm, The wavelength range of the said infrared light region is 750 nm to 1400 nm, The plant discrimination apparatus of Additional remark 1 or 2 characterized by the above-mentioned.

(Additional remark 4) The wavelength range contained in the said distribution spectrum data is 400 nm to 1400 nm, The plant discrimination apparatus of Additional remark 1, 2 or 3 characterized by the above-mentioned.

(Appendix 5) A plant discrimination method executed by a computer,
For each pixel of the image, referring to spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region,
Based on the distribution spectrum data, the set of the first median value of the light intensity of the visible light region corresponding to a certain pixel on the screen and the second median value of the light intensity of the infrared light region, A plant discrimination method comprising: performing a process of excluding a certain pixel from a plant discrimination target when a value obtained by multiplying a first median by a predetermined coefficient is smaller than the second median.

(Additional remark 6) The said predetermined coefficient is a value from 1.2 to 1.6 or a value from 1.3 to 1.4, The plant discrimination method of Additional remark 5 characterized by the above-mentioned.

(Supplementary note 7) The plant discrimination method according to supplementary note 5 or 6, wherein a wavelength range of the visible light region is 495 nm to 570 nm, and a wavelength range of the infrared light region is 750 nm to 1400 nm.

(Additional remark 8) The wavelength range contained in the said distribution spectrum data is 400 nm to 1400 nm, The plant discrimination method of Additional remark 5, 6 or 7 characterized by the above-mentioned.

(Appendix 9)
For each pixel of the image, referring to spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region,
Based on the distribution spectrum data, the set of the first median value of the light intensity of the visible light region corresponding to a certain pixel on the screen and the second median value of the light intensity of the infrared light region, A plant discrimination program that executes a process of excluding a certain pixel from a plant discrimination target when a value obtained by multiplying a first constant by a predetermined constant is smaller than the second median.

(Supplementary note 10) The plant discrimination program according to supplementary note 9, wherein the predetermined coefficient is a value from 1.2 to 1.6 or a value from 1.3 to 1.4.

(Supplementary note 11) The plant discrimination program according to supplementary note 9 or 10, wherein the visible light region has a wavelength range of 495 nm to 570 nm, and the infrared light region has a wavelength range of 750 nm to 1400 nm.

(Additional remark 12) The wavelength range contained in the said distribution spectrum data is 400 nm to 1400 nm, The plant discrimination program of Additional remark 9, 10 or 11 characterized by the above-mentioned.

100 Plant discrimination device 152 Exclusion section

Claims (4)

The light in the visible light region corresponding to a certain pixel on the image based on spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region for each pixel of the image For a set of a first median value of intensity and a second median value of light intensity of the infrared light region, a value obtained by multiplying the first median value by a predetermined coefficient is smaller than the second median value. An excluding unit for excluding the certain pixel from the discrimination target of the plant ,
The predetermined coefficient is a value from 1.2 to 1.6 or a value from 1.3 to 1.4;
The wavelength range of the visible light region is 495 nm to 570 nm, and the wavelength range of the infrared light region is 750 nm to 1400 nm . Wavelength range included in the partial light spectral data, plant determination device according to claim 1, characterized in that the 1400nm from 400 nm. A plant discrimination method executed by a computer,
For each pixel of the image, referring to spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region,
Based on the partial light spectral data, the pair of the second central value of the light intensity of the first central value and the infrared region of the light intensity of the visible light region corresponding to the screen is on the pixels, When the value obtained by multiplying the first median by a predetermined coefficient is smaller than the second median, a process of excluding the certain pixel from the plant discrimination target is performed.
The predetermined coefficient is a value from 1.2 to 1.6 or a value from 1.3 to 1.4;
The wavelength range of the visible light region is 495 nm to 570 nm, and the wavelength range of the infrared light region is 750 nm to 1400 nm . On the computer,
For each pixel of the image, referring to spectral spectrum data including light intensity information in the visible light region and light intensity information in the infrared light region,
Based on the partial light spectral data, the pair of the second central value of the light intensity of the first central value and the infrared region of the light intensity of the visible light region corresponding to the screen is on the pixels, When a value obtained by multiplying the first median value by a predetermined coefficient is smaller than the second median value, the process of excluding the certain pixel from the plant discrimination target is executed,
The predetermined coefficient is a value from 1.2 to 1.6 or a value from 1.3 to 1.4;
The wavelength range of the visible light region is 495 nm to 570 nm, and the wavelength range of the infrared light region is 750 nm to 1400 nm .

Images