JP2023151082A - Detection device, detection method and control program - Google Patents

Abstract

To provide a detection device, detection method and control program which can highly accurately detect a withered tree region where a withered tree exists.SOLUTION: A detection device comprises: an image acquisition unit which acquires an optical image obtained by imaging a target region from the sky; an index calculation unit which calculates a vegetation index for each of a plurality of pixels included in the optical image; a setting unit which sets a plurality of section regions in the optical image; a dispersion degree calculation unit which calculates a dispersion degree of the vegetation index of each pixel included in each section region for each of the plurality of section regions; a detection unit which detects a withered tree region where a withered tree exists from the plurality of section regions on the basis of the dispersion degree; and an output unit which outputs information about a withered tree region.SELECTED DRAWING: Figure 2

Description

The present invention relates to a detection device, a detection method, and a control program.

In recent years, the number of trees subject to inspection has been increasing due to the progress of growing larger and older trees on premises, as well as an increase in the number of unmanaged trees, and there is a need to improve the efficiency of inspection work.

Patent Document 1 describes a tree health determination method that determines tree health according to tree species by combining tree species estimated by analyzing satellite images and a normalized difference vegetation index (NDVI). The method is described.

Japanese Patent Application Publication No. 2019-144607

In tree inspection work, it is required to detect with high accuracy the dead tree area where dead trees are present.

An object of the present invention is to provide a detection device, a detection method, and a control program that can detect with high accuracy a dead tree area where dead trees exist.

A detection device according to one aspect of the present invention includes: an image acquisition unit that acquires an optical image of a target area photographed from above; an index calculation unit that calculates a vegetation index for each of a plurality of pixels included in the optical image; a setting section that sets a plurality of segmented areas within the area; a dispersion degree calculation section that calculates the degree of variation in the vegetation index of each pixel included in each segmented area for each of the plurality of segmented areas; The present invention is characterized by having a detection unit that detects a dead tree area in which a dead tree exists from among the divided areas, and an output unit that outputs information regarding the dead tree area.

In the detection device according to one aspect of the present invention, the detection unit is configured such that the average value of the vegetation index of each pixel included in each segmented region among the plurality of segmented regions is smaller than a first threshold value, and It is preferable that a segmented area in which the standard deviation of the vegetation index of each pixel is larger than a second threshold value be detected as a dead tree area.

In the detection device according to one aspect of the present invention, the index calculation unit calculates a GSI index for each of the plurality of pixels included in the optical image, and calculates a vegetation area from the optical image based on the difference between the vegetation index and the GSI index. It is preferable that the image forming apparatus further includes an extracting section for extracting data, and that the setting section sets a plurality of segmented regions within the vegetation region.

A detection method according to one aspect of the present invention uses a computer to acquire an optical image of a target area photographed from above, calculates a vegetation index for each of a plurality of pixels included in the optical image, and divides a target area into a plurality of divisions within the optical image. The areas are set, and the degree of variation in the vegetation index of each pixel included in each divided area is calculated for each divided area, and based on the degree of variation, dead trees where dead trees exist are selected from among the multiple divided areas. It is characterized by detecting tree areas and outputting information regarding damaged wood areas.

A control program according to one aspect of the present invention is a control program for a computer having an output unit, which acquires an optical image of a target area taken from above, and calculates a vegetation index for each of a plurality of pixels included in the optical image. Then, multiple segmented areas are set in the optical image, and the degree of variation in the vegetation index of each pixel included in each segmented area is calculated for each of the multiple segmented areas. The present invention is characterized in that the detecting device detects a dead tree area in which a dead tree exists, and outputs information regarding the damaged tree area from an output unit.

According to the present invention, the detection device, the detection method, and the control program can detect with high precision a dead tree area where dead trees exist.

FIG. 1 is a diagram showing an example of a schematic configuration of a detection device. It is a flow chart which shows an example of dead tree area detection processing. FIG. 3 is a diagram showing an example of an RGB image. It is a figure which shows an example of a red band image. It is a figure which shows an example of a green band image. FIG. 3 is a diagram showing an example of a blue band image. It is a figure showing an example of a near-infrared band image. It is a figure which shows an example of a vegetation index image. FIG. 3 is a diagram showing an example of a GSI index image. FIG. 3 is a schematic diagram for explaining a vegetation area. FIG. 3 is a schematic diagram for explaining divided areas. It is a figure which shows an example of a variation degree image. It is a figure showing an example of a representative value image.

Hereinafter, various embodiments of the present invention will be described with reference to the drawings. It should be noted that the technical scope of the present invention is not limited to these embodiments, but extends to the invention described in the claims and equivalents thereof.

FIG. 1 is a diagram showing an example of a schematic configuration of a detection device 1. As shown in FIG. The detection device 1 is an example of a computer, and detects a dead tree area where a dead tree exists. The detection device 1 includes a storage section 11, a communication section 12, a display section 13, an operation section 14, and a processing section 15.

The storage unit 11 stores programs or data. The storage unit 11 includes, for example, a semiconductor memory device. The storage unit 11 stores operating system programs, driver programs, application programs, data, etc. used in processing by the processing unit 15. The program is stored in a computer-readable, non-temporary, portable storage medium such as a CD (Compact Disc)-ROM (Read Only Memory) or a DVD (Digital Versatile Disc)-ROM using a known setup program or the like. Installed on 11.

The communication unit 12 is an example of an output unit. The communication unit 12 enables the detection device 1 to communicate with other devices. The communication unit 12 includes a communication interface circuit. The communication interface circuit included in the communication unit 12 is a communication interface circuit such as a wired LAN (Local Area Network) or a wireless LAN. The communication unit 12 receives data from another device and supplies it to the processing unit 15, and also transmits the data supplied from the processing unit 15 to the other device.

The display section 13 is an example of an output section. The display unit 13 displays images. The display unit 13 includes, for example, a liquid crystal display or an organic EL (Electro-Luminescence) display. The display unit 13 displays images based on display data supplied from the processing unit 15.

The operation unit 14 accepts a user's input operation on the detection device 1 . The operation unit 14 includes, for example, a keypad, a keyboard, or a mouse. The operation unit 14 may include a touch panel integrated with the display unit 13. The operation unit 14 generates a signal according to the user's input operation and supplies it to the processing unit 15.

The processing unit 15 is a device that centrally controls the operation of the detection device 1, and includes one or more processors and their peripheral circuits. The processing unit 15 includes, for example, a CPU (Central Processing Unit). The processing unit 15 may include a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), and the like. The processing unit 15 controls the operation of each component so that various processes of the detection device 1 are executed in appropriate procedures based on the program stored in the storage unit 11 and inputs from the communication unit 12 and the operation unit 14. It also controls and executes various processes.

The processing unit 15 includes an image acquisition unit 151, an index calculation unit 152, an extraction unit 153, a setting unit 154, a variation degree calculation unit 155, a detection unit 156, and an output control unit 157 as functional blocks. Each of these units is a functional module realized by a program executed by the processing unit 15. Each of these parts may be implemented in the detection device 1 as firmware.

FIG. 2 is a flowchart illustrating an example of dead tree area detection processing. The dead tree area detection process is mainly executed by the processing unit 15 in cooperation with each element of the detection device 1 based on a program stored in the storage unit 11 in advance.

First, the image acquisition unit 151 acquires an optical image of the target area taken from above from an external information processing device (for example, an information processing device mounted on an optical satellite) via the communication unit 12 (step S101). . The target area is an area including trees such as a forest. The detection device 1 inspects trees included in a target area, and detects a dead tree area in which a dead tree, which is a dead tree, exists from the target area.

The optical image is a Pleiades optical satellite image or the like. The optical image shows the intensity of sunlight reflected light (electromagnetic waves) and emitted light (electromagnetic waves) from each position at each position within the target area, detected by the optical sensor on board the optical satellite, corresponding to each position. This is an image with the pixel value of each pixel. The optical image includes a plurality of images generated for each wavelength of a predetermined width (band) among the wavelengths of electromagnetic waves detected by an optical sensor mounted on an optical satellite.

The optical image includes a red band image, a green band image, a blue band image, and a near-infrared band image. A red band image is an image in which the pixel value of each pixel is the intensity of light in a band that includes the wavelength of red light. A green band image is an image in which the pixel value of each pixel is the intensity of light in a band that includes the wavelength of green light. A blue band image is an image in which the pixel value of each pixel is the intensity of light in a band that includes the wavelength of blue light. A near-infrared band image is an image in which the pixel value of each pixel is the intensity of light in a band including the wavelength of near-infrared light.

The optical image further includes an RGB image. In an RGB image, the pixel value of each pixel is divided into the pixel value (16 bits) of the corresponding pixel in the red band image, the pixel value (16 bits) of the corresponding pixel in the green band image, and the pixel value (16 bits) of the corresponding pixel in the blue band image. This is a color image with a 48-bit color value consisting of a value (16 bits). Here, the number of bits of the pixel value is an example, and pixels having a pixel value of 8 bits or 12 bits may be used depending on the specifications of the satellite.

FIG. 3A shows an example of an RGB image 300 of a predetermined target area 301 captured from above.

As shown in FIG. 3A, a target area 301 included in an RGB image 300 includes a vegetation area 302 that includes trees and a non-vegetation area 303 that does not include trees. The vegetation area 302 also includes a dead tree area 304 where dead trees exist and a non-dead tree area 305 where no dead trees exist. The non-vegetated area 303 includes a soil area 306 where soil is exposed on the ground surface like a bare field area, and a non-soil area 307 where concrete or the like is present and soil is not exposed on the ground surface.

FIG. 3B shows an example of a red band image 310 captured from above the target area 301.

In the example shown in FIG. 3B, pixels corresponding to positions where the intensity of emitted or reflected red light is greater are shown with higher luminance (closer to white), and pixels corresponding to positions where the intensity of emitted or reflected red light is smaller. The brighter the image, the lower the brightness (closer to black).

FIG. 4A shows an example of a green band image 400 obtained by capturing an image of the target region 301 from above.

In the example shown in FIG. 4A, pixels corresponding to positions where the intensity of emitted or reflected green light is greater are shown with higher luminance (closer to white), and pixels corresponding to positions where the intensity of emitted or reflected green light is smaller. The brighter the image, the lower the brightness (closer to black).

FIG. 4B shows an example of a blue band image 410 captured from above the target area 301.

In the example shown in FIG. 4B, pixels corresponding to positions where the intensity of emitted or reflected blue light is higher are shown with higher luminance (closer to white), and pixels corresponding to positions where the intensity of emitted or reflected blue light is lower. The brighter the image, the lower the brightness (closer to black).

FIG. 5A shows an example of a near-infrared band image 500 obtained by capturing an image of the target region 301 from above.

In the example shown in FIG. 5A, the pixel corresponding to the position where the intensity of near-infrared light emitted or reflected is higher is shown with higher luminance (closer to white), and the position where the intensity of near-infrared light emitted or reflected is lower. A pixel corresponding to a pixel is shown with lower luminance (closer to black).

Next, the index calculating unit 152 calculates a vegetation index from the optical image for each of a plurality of pixels included in the optical image (step S102). The vegetation index is an index for understanding the status of vegetation in a target area, and indicates the amount or vitality of plants. The index calculation unit 152 calculates a wide dynamic range vegetation index (WDRVI) as a vegetation index. WDRVI has a higher value as the amount or vigor of plants is greater, and a lower value as the amount or vigor of plants is smaller. The index calculation unit 152 calculates the vegetation index of each pixel from the following formula (1) based on the red band image and the near-infrared band image.
WDRVI=(αIR−R)/(αIR+R)…(1)

Here, α is a correction coefficient, and is set to a value within a range of, for example, 0.1 or more and 0.2 or less. IR is the pixel value of a corresponding pixel in the near-infrared band image, that is, the intensity of near-infrared light at the position corresponding to that pixel. R is the pixel value of the corresponding pixel in the red band image, ie the intensity of the red light at the position corresponding to that pixel. WDRVI has a value within a range of -1.0 or more and +1.0 or less. It is known that plants have a high reflectance of near-infrared light and a low reflectance of red light due to the action of chlorophyll. It is also known that when plants become weakened, their chlorophyll activity slows down, and their near-infrared light reflectance decreases. Therefore, the detection device 1 can improve the detection accuracy of the vegetation area by using the WDRVI calculated by the above equation (1) as a vegetation index.

Note that the index calculation unit 152 may calculate a normalized difference vegetation index (NDVI) as a vegetation index instead of WDRVI. In that case, the index calculation unit 152 calculates the vegetation index of each pixel from the following formula (2) based on the red band image and the near-infrared band image.
NDVI=(IR-R)/(IR+R)...(2)

Note that WDRVI is a vegetation index that has been improved to alleviate the problem of saturation of the NDVI value in areas where the Leaf Area Index (LAI) is large. LAI is an index representing the amount of leaves in a plant community, and the more leaves there are, the larger the LAI value is. By using WDRVI as a vegetation index, the detection device 1 can further improve the detection accuracy of a vegetation area, compared to using NDVI as a vegetation index.

The index calculation unit 152 generates a vegetation index image using the pixel value of each pixel as a vegetation index.

FIG. 5B shows an example of a vegetation index image 510 generated from the red band image 310 and the near-infrared band image 500 using WDRVI as a vegetation index.

In the example shown in FIG. 5B, a pixel with a higher vegetation index is shown with higher brightness (closer to white), and a pixel with a lower vegetation index is shown with lower brightness (closer to black). As shown in FIG. 5B, among the regions 511 corresponding to the target region 301, a region 512 corresponding to the vegetation region 302 has a high vegetation index, and a region 513 corresponding to the non-vegetation region 303 has a low vegetation index. Therefore, the detection device 1 can accurately identify a vegetation area and a non-vegetation area by using WDRVI as a vegetation index.

Note that among the regions 512 corresponding to the vegetation regions 302, the vegetation index is high overall in the region 515 corresponding to the non-dead tree region 305, and the vegetation index is high in some regions in the region 514 corresponding to the dead tree region 304. Slightly lower. Further, among the regions 513 corresponding to the non-vegetated regions 303, the vegetation index is sufficiently low in the region 517 corresponding to the non-soil region 307, but the vegetation index is slightly high in the region 516 corresponding to the soil region 306.

Next, the index calculating unit 152 calculates a GSI index (Grain Size Index) from the optical image for each of the plurality of pixels included in the optical image (step S103). The GSI index is also called a grain size index and is an index for detecting bare ground areas. The GSI index has a higher value as the position corresponding to the pixel is more likely to be included in a bare ground area. The index calculation unit 152 calculates the GSI index of each pixel from the following formula (3) based on the red band image, green band image, and blue band image.
GSI index = (R-B)/(R+G+B)...(3)

Here, G is the pixel value of the corresponding pixel in the green band image, ie, the intensity of green light at the position corresponding to that pixel. B is the pixel value of the corresponding pixel in the blue band image, ie, the intensity of blue light at the position corresponding to that pixel. Like WDRVI, the GSI index has a value within a range of -1.0 or more and +1.0 or less.

The index calculation unit 152 generates a GSI index image using the pixel value of each pixel as a GSI index.

FIG. 6A shows an example of a GSI index image 600 generated from a red band image 310, a green band image 400, and a blue band image 410.

In the example shown in FIG. 6A, a pixel with a higher GSI index is shown with higher brightness (closer to white), and a pixel with a lower GSI index is shown with lower brightness (closer to black). As shown in FIG. 7A, in the region 601 corresponding to the target region 301, the region 606 corresponding to the soil region 306 is compared with the region 602 corresponding to the vegetation region 302 and the region 607 corresponding to the non-soil region 307. , the GSI index is high. Therefore, the detection device 1 can accurately identify soil areas and other areas by using the GSI index.

Next, the index calculating unit 152 calculates the difference between the vegetation index and the GSI index for each of the plurality of pixels included in the optical image (step S104). The index calculating unit 152 calculates the difference value of each pixel by subtracting the pixel value of the corresponding pixel of the GSI index image from the pixel value of the corresponding pixel of the vegetation index image. Generate as a difference index image with the index image.

As described above, the vegetation index is an index for grasping the state of vegetation in a target area, and has a high value in a vegetated area and a low value in a non-vegetated area. However, among the non-vegetated areas, the soil area may include grassland or farmland, and in that case, the vegetation index has a high value even in the soil area. The GSI index has a higher value in the area corresponding to the soil area than in other areas. Therefore, the difference obtained by subtracting the GSI index from the vegetation index maintains a high value close to the vegetation index in the vegetation area, and has a sufficiently low value with respect to the vegetation index in the soil area. Note that in the non-soil area, since the vegetation index is low, the difference obtained by subtracting the GSI index from the vegetation index has a low value. The index calculation unit 152 can reduce the possibility of extracting grassland or farmland as a vegetation area by correcting the vegetation index using the GSI index that indicates the likelihood of a bare land area.

Next, the extraction unit 153 performs binarization processing on the difference index image generated by the index calculation unit 152 (step S105). The extraction unit 153 determines whether the pixel value of each pixel included in the difference index image is equal to or greater than a threshold value. The threshold value is determined, for example, by Otsu's binarization. Note that the threshold value may be set to the average value, median value, or the like of the pixel values of each pixel included in the difference index image. Further, the threshold value may be set to a fixed value in advance. The extraction unit 153 sets the pixel value of each pixel as a valid value when the pixel value of the corresponding pixel in the difference index image is equal to or greater than the threshold value, and sets the pixel value of each pixel as a valid value when the pixel value of the corresponding pixel in the difference index image is less than the threshold value. Generate a binary image with invalid values.

Next, the extraction unit 153 extracts a vegetation area from the optical image based on the difference between the vegetation index calculated by the index calculation unit 152 and the GSI index (step S106). The extraction unit 153 performs processing such as labeling on the binary image, and extracts a region made up of mutually adjacent valid pixels as a valid region. The extraction unit 153 generates data indicating the coordinates of the extracted effective area as vegetation area polygon data, and extracts an area corresponding to the vegetation area polygon data as the vegetation area within the optical image.

FIG. 6B is a schematic diagram for explaining the vegetation area.

The image 610 shown in FIG. 6B is the same image as the RGB image 300 shown in FIG. 3A. In the image 610, an area 612 indicates an area corresponding to the vegetation area 302 extracted by the extraction unit 153. As shown in FIG. 6B, each area corresponding to the vegetation area 302 is extracted as a vegetation area 612 by the extraction unit 153. In this way, the detection device 1 can detect a vegetation area with high precision by using the difference between the vegetation index and the GSI index.

Moreover, since the process is executed only for this vegetation area from now on, the detection device 1 can speed up the dead tree area detection process. Note that, if the land use situation has been obtained in advance, for example from a high-resolution land use land cover map of the Japan area provided by JAXA, the extraction unit 153 extracts the vegetation area based on the use situation. You can.

Next, the setting unit 154 sets a plurality of segmented areas within the optical image (step S107). The setting unit 154 sets a plurality of segmented areas by dividing the vegetation area extracted by the extraction unit 153 within the optical image into a plurality of areas. The setting unit 154 sets each area including three or more pixels in the horizontal direction and the vertical direction in the optical image as a segmented area. Thereby, the detection device 1 can accurately calculate the degree of variation and representative value of the vegetation index in each segmented area in the process described below. The setting unit 154 sets a square area including three pixels each in the horizontal direction and the vertical direction within the optical image as a segmented area.

Note that the setting unit 154 may set rectangular areas with different sizes in the horizontal direction and vertical direction as the segmented areas. Furthermore, the setting unit 154 may set an area having another shape, such as a circular area or a triangular area, as the segmented area.

FIG. 7A is a schematic diagram for explaining the segmented areas.

Image 700 shown in FIG. 7A is the same image as image 610 shown in FIG. 6B. In the image 700, an area 701 indicates an area corresponding to the segmented area set by the setting unit 154. As shown in FIG. 7A, a plurality of segmented regions 701 are set by dividing a region 612 corresponding to the vegetation region 302 extracted by the extraction unit 153 into a plurality of rectangular regions.

Next, the variation degree calculation unit 155 calculates the degree of variation in the vegetation index of each pixel included in each segmented area for each of the plurality of segmented areas (step S108). The degree of variation calculation unit 155 calculates the standard deviation of the vegetation index of each pixel included in each segmented area as the degree of variation. Note that the degree-of-dispersion calculation unit 155 may calculate other indicators such as dispersion and half-width (the width of the mountain where the peak value of the distribution of the vegetation index is halved) as the degree of dispersion. The dispersion degree calculation unit 155 generates a dispersion degree image in which the pixel value of each pixel is the dispersion degree of the segmented area including the pixel corresponding to each pixel.

FIG. 7B shows an example of a variation degree image 710 generated from the red band image 310 and the near-infrared band image 500, using the standard deviation as the variation degree. Note that in the variation degree image 710 shown in FIG. 7B, in order to improve visibility, the corresponding areas of the RGB image 300 are shown for areas other than the area 712 corresponding to the vegetation area 302.

In the example shown in FIG. 7B, pixels with a greater degree of variation are shown with higher luminance (closer to white), and pixels with a smaller degree of variation are indicated with lower luminance (closer to black). As shown in FIG. 7B, in a region 712 corresponding to the vegetation region 302, a region 714 corresponding to the dead tree region 304 has a greater degree of variation than a region 715 corresponding to the non-dead tree region 305.

As described above, among the regions 712 corresponding to the vegetation regions 302, the vegetation index is large overall in the region 715 corresponding to the non-dead tree region 305, and in some regions in the region 714 corresponding to the dead tree region 304. The vegetation index is slightly smaller. However, the difference in the vegetation index value itself between the region 715 corresponding to the non-dead tree region 305 and the region 714 corresponding to the dead tree region 304 is small, and based on the vegetation index value itself, It is difficult to distinguish between the dead tree area 305 and the dead tree area 305. On the other hand, the detection device 1 can accurately identify dead tree areas and non-dead tree areas by using the degree of variation in the vegetation index.

Next, the variation degree calculation unit 155 calculates, for each of the plurality of divided regions, a representative value of the vegetation index of each pixel included in each divided region (step S109). The degree of variation calculation unit 155 calculates the average value of the vegetation index of each pixel included in each segmented area as a representative value.

Note that the variation degree calculation unit 155 may calculate other indicators such as a maximum value, a minimum value, a median value, etc. as a representative value. The degree-of-dispersion calculation unit 155 generates a representative value image in which the pixel value of each pixel is a representative value of a segmented area in which the pixel corresponding to each pixel is included.

FIG. 8 shows an example of a representative value image 800 generated from the red band image 310 and the near-infrared band image 500, using the average value as the representative value. Note that in the representative value image 800 shown in FIG. 8, in order to improve visibility, the corresponding areas of the RGB image 300 are shown for areas other than the area 802 corresponding to the vegetation area 302.

In the example shown in FIG. 8, a pixel with a smaller representative value is shown with higher luminance (nearly white), and a pixel with a larger representative value is shown with lower luminance (closer to black). As shown in FIG. 8, in the region 802 corresponding to the vegetation region 302, the representative value tends to be smaller in the region 804 corresponding to the dead tree region 304 compared to the region 805 corresponding to the non-dead tree region 305. be. The detection device 1 can identify a dead tree area and a non-dead tree area with higher accuracy by using the representative value of the vegetation index in addition to the degree of variation in the vegetation index.

Next, the detection unit 156 detects a dead tree area where a dead tree exists from among the plurality of segmented areas based on the degree of variation and the representative value calculated by the degree of variation calculation unit 155 (step S110). The detection unit 156 determines that among the plurality of divided regions, the representative value of the vegetation index of each pixel included in each divided region is smaller than the first threshold, and the degree of variation in the vegetation index of each pixel included in each divided region is A segmented area larger than the second threshold is detected as a dead tree area.

For example, the detection unit 156 determines that the average value of the vegetation index of each pixel included in each divided area among the plurality of divided areas is smaller than the first threshold value, and that the standard of the vegetation index of each pixel included in each divided area is A segmented area in which the deviation is larger than the second threshold is detected as a dead tree area.

The first threshold value is, for example, the average value or maximum value of the representative values calculated for the dead tree area included in various optical images and the average value or the maximum value of the representative values calculated for the non-dead tree area in a preliminary experiment. Set to a value between the minimum value. The second threshold is, for example, the average value or minimum value of the degree of variation calculated for dead tree areas included in various optical images in a preliminary experiment, and the average value or minimum value of the degree of variation calculated for non-dead tree areas in a prior experiment. Set to a value between the maximum value.

On the other hand, the detection unit 156 detects, among the plurality of divided regions, a divided region in which the representative value of the vegetation index of each pixel included in each divided region is greater than or equal to the first threshold value, and the vegetation index of each pixel included in each divided region. A segmented area in which the degree of variation in the index is less than or equal to the second threshold is detected as a non-dead tree area.

Normally, in an area that includes multiple trees, it is unlikely that all the trees will die all at once, but it is highly likely that the multiple trees will die one after another. Therefore, a divided area that includes dead trees is likely to also include healthy trees, and the variation in vegetation indices tends to be larger than a divided area that includes only healthy trees. Therefore, the detection device 1 can appropriately detect the dead tree area through image processing by utilizing the degree of variation in the vegetation index.

Further, in an area having a certain size, the vegetation index becomes small for dead trees, but the vegetation index becomes large for healthy trees. By using the representative value of the vegetation index in a segmented area having a certain size, the detection device 1 can accurately detect a segmented area in which dead trees exist only in part as a dead tree area.

For example, even if the representative value of the vegetation index is large, if the degree of variation in the vegetation index is large, there is a high possibility that dead trees and healthy trees are mixed. The detection device 1 can improve the detection accuracy of the dead tree area by using both the degree of variation and the representative value.

Note that the detection unit 156 may calculate an evaluation value from the variation degree and representative value of the vegetation index for each of the plurality of divided areas, and detect the dead tree area based on the calculated evaluation value. The detection unit 156 calculates the evaluation value so that the evaluation value increases as the degree of variation in the vegetation index increases, and increases as the representative value of the vegetation index decreases.

The detection unit 156 detects, among the plurality of segmented regions, a segmented region whose evaluation value is larger than a threshold value as a dead tree region, and detects a segmented region whose evaluation value is less than or equal to the threshold value as a non-dead tree region. The threshold value is, for example, the average value or minimum value of the evaluation values calculated for the dead tree area included in various optical images and the average value or maximum value of the evaluation values calculated for the non-dead tree area in a preliminary experiment. is set to a value between . In this case as well, the detection unit 156 can detect the dead tree area with high accuracy.

Furthermore, the detection unit 156 changes a second threshold for comparing the degree of variation in the vegetation index based on the representative value of the vegetation index for each of the plurality of divided regions, so that the degree of variation in the vegetation index is lower than the second threshold. It may be determined whether each divided area is a dead tree area or not depending on whether the divided area is large or not. In that case, the detection unit 156 makes the second threshold smaller as the representative value of the vegetation index is smaller, making it easier to determine that each segmented area is a dead tree area.

Note that the detection unit 156 changes the first threshold value for comparison with the representative value of the vegetation index based on the degree of variation in the vegetation index for each of the plurality of divided regions, so that the representative value of the vegetation index is lower than the first threshold value. It may be determined whether each divided area is a dead tree area or not depending on whether the area is small or not. In this case, the detection unit 156 increases the first threshold value as the degree of variation in the vegetation index increases, making it easier to determine that each segmented area is a dead tree area. In these cases as well, the detection unit 156 can detect the dead tree area with high accuracy.

Note that the detection unit 156 classifies the segmented areas into groups using a known numerical classification technique, and determines for each segmented area group classified into groups whether or not the segmented area group is a dead tree area. You may. In that case, the degree of variation calculation unit 155 classifies divided regions that are adjacent to each other and have similar representative values of vegetation indices (the difference between the representative values is equal to or less than a predetermined value) into the same group, for example.

The degree of variation calculation unit 155 sets a predetermined value using, for example, natural classification technology so that adjacent segmented areas in which the difference in representative values of vegetation indices is relatively large are classified into different groups. You can. The variation degree calculation unit 155 calculates the variation degree and representative value of the vegetation index for each classified area group. The detection unit 156 determines that among the plurality of segmented area groups, the representative value of the vegetation index of each pixel included in each segmented area group is smaller than the first threshold value, and that the representative value of the vegetation index of each pixel included in each segmented area group is A group of segmented areas with a degree of variation greater than a second threshold are collectively detected as a dead tree area.

Since the detection unit 156 can detect dead tree areas all at once, it is possible to reduce the processing time required for dead tree area detection processing. Further, the detection unit 156 can reduce failure to detect a dead tree area or error in detecting a dead tree area due to the influence of noise or the like in the image.

Next, the output control unit 157 outputs information regarding the dead tree area (step S111), and ends the series of dead tree area detection processing. The output control unit 157 outputs information regarding the dead tree area detected by the detection unit 156 by displaying it on the display unit 13. The output control unit 157 may output information regarding the dead tree area by transmitting it to an external information processing device via the communication unit 12. The output control unit 157 highlights the dead tree area by displaying it in a different color from other areas on the RGB image or by surrounding it with a thick line, for example, and generates an image in which the area is displayed distinguishably from other areas. and outputs it as information regarding the dead tree area. The output control unit 157 may output the latitude or longitude indicating the dead tree area as information regarding the dead tree area.

Note that the process in step S103 may be omitted, and in step S106, the extraction unit 153 may extract the vegetation area using only the vegetation index image without using the GSI index image. Further, in step S106, the extraction unit 153 may extract the vegetation area using only the GSI index image without using the vegetation index image. Further, the processing in steps S103 to S106 may be omitted, and in step S107, the setting unit 154 may set a plurality of segmented areas in the entire area of the optical image. Alternatively, the process in step S109 may be omitted, and in step S110, the setting unit 154 may detect the dead tree area using only the degree of variation in the vegetation index without using the representative value of the vegetation index.

As described in detail above, the detection device 1 detects a dead tree area based on the degree of variation in the vegetation index in each divided area. As a result, the detection device 1 can detect as a dead tree area a divided area in which only some of the trees are dead trees, or a divided area in which trees that have not completely withered but are weakened are included. . Therefore, the detection device 1 can detect with high accuracy a dead tree area where dead trees exist.

Further, since the detection device 1 automatically detects the dead tree area through image processing, the user can efficiently perform tree inspection work. Therefore, the detection device 1 can improve convenience for the user.

It should be understood that those skilled in the art can make various changes, substitutions, and modifications thereto without departing from the spirit and scope of the invention. For example, the processing of each part described above may be executed in a different order as appropriate within the scope of the present invention. Furthermore, the embodiments and modifications described above may be implemented in appropriate combinations within the scope of the present invention.

1 Detection device 151 Image acquisition section 152 Index calculation section 153 Extraction section 154 Setting section 155 Dispersion degree calculation section 156 Detection section 157 Output control section

Claims (5)

an image acquisition unit that acquires an optical image of the target area taken from above;
an index calculation unit that calculates a vegetation index for each of a plurality of pixels included in the optical image;
a setting unit that sets a plurality of segmented areas within the optical image;
a variation degree calculation unit that calculates, for each of the plurality of segmented areas, a degree of variation in the vegetation index of each pixel included in each segmented area;
a detection unit that detects a dead tree area in which a dead tree exists from among the plurality of divided areas based on the degree of variation;
an output unit that outputs information regarding the dead tree area;
A detection device comprising: The detection unit is configured such that an average value of the vegetation index of each pixel included in each segmented region among the plurality of segmented regions is smaller than a first threshold value, and an average value of the vegetation index of each pixel included in each segmented region is The detection device according to claim 1, wherein a segmented area whose standard deviation is larger than a second threshold is detected as the dead tree area. The index calculation unit calculates a GSI index for each of a plurality of pixels included in the optical image,
further comprising an extraction unit that extracts a vegetation area from the optical image based on a difference between the vegetation index and the GSI index,
The detection device according to claim 1 or 2, wherein the setting unit sets the plurality of divided areas within the vegetation area. By computer,
Obtain an optical image of the target area taken from above,
calculating a vegetation index for each of a plurality of pixels included in the optical image;
setting a plurality of segmented areas within the optical image;
For each of the plurality of divided areas, calculate the degree of variation in the vegetation index of each pixel included in each divided area,
detecting a dead tree area in which a dead tree exists from among the plurality of divided areas based on the degree of variation;
outputting information regarding the dead tree area from an output unit;
A detection method characterized by: A control program for a computer having an output section,
Obtain an optical image of the target area taken from above,
calculating a vegetation index for each of a plurality of pixels included in the optical image;
setting a plurality of segmented areas within the optical image;
For each of the plurality of divided areas, calculate the degree of variation in the vegetation index of each pixel included in each divided area,
detecting a dead tree area in which a dead tree exists from among the plurality of divided areas based on the degree of variation;
outputting information regarding the dead tree area from the output unit;
A control program characterized by causing a detection device to perform the following.

Images