JPH09257577A - Identification method of broad-leaved weeds in a meadow using spectral reflectance - Google Patents

Abstract

(57) 【Abstract】 PROBLEM TO BE SOLVED: By continuously measuring the spectral reflectance of the surface of a grassland, a grass having a relatively small leaf width and a broad-leaved weed having a much wider leaf width than that of the grass are measured. To identify. SOLUTION: Rumex japonicus X (broad leaf weed) to be identified
The optimum measurement area A of the spectral reflectance measuring device based on the average width of each leaf of the and orchard grass OG (grass) and the optimum moving speed V of the measuring device for the measuring area A are determined in advance. If a plurality of objects to be identified are included in the moving measurement area A, the combined spectral reflectances thereof are calculated, and the spectral reflectances of the objects to be identified on the meadow at a specific wavelength are continuously calculated. And the measured spectral reflectances are included in the range of spectral reflectances that are regarded as preset broadleaf weeds, and the identified object is treated as a broadleaf weed only when it is included a predetermined number of times consecutively. Is to determine that.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention moves a spectral reflectance measuring instrument in a meadow to continuously measure the spectral reflectance of an object to be identified in a field scene, so that a grass having a relatively small leaf width can be obtained. , It relates to a method of distinguishing from broad-leaved weeds whose leaves are much wider than this.

[0002]

2. Description of the Related Art The applicant of the present invention has made an object to be identified by relatively moving the object to be identified and a measuring head for measuring the spectral radiance of the object to be identified by changing the distance therebetween. A technique for optically identifying the object to be identified by measuring the spectral solid angle reflectance (hereinafter, simply referred to as “spectral reflectance”) of JP-A-7-11123 has been completed.
1, filed 7-11132.

When this technique is used to detect only weeds in a meadow and selectively spray a herbicide thereto, the spectral reflectance of the target weed and that of the pasture are almost the same. Then, it is difficult to identify the spot reflectances only by measuring the spectral reflectances of the both. The present inventor has conducted repeated research to identify two or more objects by using the spectral reflectances, and as a result, even if the spectral reflectances measured when two different objects are stopped are almost the same, the two objects are In the case where the widths of the two are significantly different, it was found that the spectral reflectances before and after that can be continuously measured to distinguish the two.

[0004]

In view of the above, the present invention is to compare the leaf width by moving the spectral reflectance measuring device in the meadow and continuously measuring the spectral reflectance. The challenge is to distinguish between small grasses and broad-leaved weeds that are much wider than this.

[0005]

DISCLOSURE OF THE INVENTION The present invention for solving this problem continuously measures the spectral reflectance of the object to be identified while moving the spectral reflectance measuring device along the surface of the meadow. And a method for distinguishing between a grass having a relatively small leaf width and a broadleaf weed having a much wider leaf width,
The optimum measurement area of the spectral reflectance measuring instrument based on the average width of each leaf of broad-leaved weeds and pastures to be identified, and the optimum moving speed of the measuring instrument for the measurement area are set in advance. In addition, when the moving measurement area includes a plurality of objects to be identified, the composite spectral reflectance thereof is calculated, and the spectral reflectance of the object to be identified of the meadow at a specific wavelength is continuously measured. Then, the measured spectral reflectance is a broadleaf weed only when it is included in a range of spectral reflectance that is regarded as a preset broadleaf weed, and only when it is included a predetermined number of times. The feature is that it is determined.

First, the optimum measurement area and measurement speed of the spectral reflectance measuring instrument for distinguishing grass and broadleaf weeds are determined in advance. If the measurement area is too small for each width of the grass and broad-leaved weeds that are the objects to be identified, the width of the leaves will be wider than the measurement area, so it is necessary to increase the measurement speed. Also, if the measurement area is too large, the synthetic spectral reflectance of grass and broadleaf weeds is detected in most parts, and both become smaller than the threshold value,
It becomes impossible to determine regardless of the measurement speed. Therefore, based on the optimum measurement area and measurement speed of the spectral reflectance measuring instrument for distinguishing between grass and broad-leaved weeds, the measuring instrument is moved, and the grass that is the identification target in the moving measurement area. If the object to be identified other than broadleaf weeds (for example, soil) is included, the composite spectral reflectances of multiple objects to be identified including grass or broadleaf weeds are detected before and after the measurement area passes through the grass or broadleaf weeds. It Therefore, when the measurement area passes before and after a grass with a narrow leaf width, the spectral reflectance of the grass is affected by the synthetic spectral reflectance, and the continuously measured spectral reflectance is only the grass. It becomes smaller than the spectral reflectance measured by spot. On the other hand, in the case of broad-leaved weeds with a large leaf width, when the measurement area passes through the center of the weeds, the broad-leaved weeds are hardly affected by the objects to be identified existing before and after the broad-leaved weeds. The original spectral reflectance of weeds is measured as is. This makes it possible to distinguish between a grass with a narrow leaf width and a broadleaf weed with a much wider leafwidth, and the measured spectral reflectance is the spectral reflectance that is regarded as a preset broadleaf weed. If the object to be identified is determined to be a broad-leaved weed only when the object is continuously included within the range of a predetermined number of times, the identification accuracy increases.

Then, the spectral reflectances at a plurality of different wavelengths as well as at a specific one wavelength are measured as described above, and the object to be identified is regarded as a broadleaf weed at any of the plurality of wavelengths. In the first, when this is discriminated as a broad-leaved weed for the first time, its identification accuracy is further enhanced.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in more detail by way of examples. First, an outline of a spectral reflectance measuring instrument will be described, and then a method for identifying broadleaf weeds in a meadow according to the present invention using this spectral reflectance measuring instrument will be described. FIG. 1 is a principle diagram of an optical identification method of an object using a spectral reflectance measuring device, and FIG. 2 is a diagram showing a plurality of detectors C.
And ₁ -C _6, an ultrasonic sensor 2, in order to explain the arrangement of the light source 3, is a view of the measuring head H from the bottom. 1 and 2, the measuring device is composed of a measuring device main body B and a measuring head H. The measuring head H is the identified object M.
For detecting the spectral radiance of the light, and by the action of the interference filter 1 provided, it is possible to respectively detect only the lights of the plurality of different set wavelengths reflected by the object M to be identified. Devices C _{1 to} C ₆ , an ultrasonic sensor 2 for detecting a distance (h) between the object M to be identified and the measuring head H, and a light source 3 including a halogen lamp.
And The detectors C _{1 to} C ₆ of the embodiment are
It consists of an optical element with a silicon photodiode. In the measuring head H shown in FIG. 1, a circular ultrasonic sensor 2 and a circular light source 3 are concentrically arranged in the center thereof, and six types of circular detectors C ₁ to C ₁ are arranged around them. C ₆
Are arranged.

Each of the detectors C _{1 to} C ₆ detects only the light of the set wavelength corresponding to the interference filter 1 provided therein, and the plurality of kinds of light are guided to the measuring device main body B via the cable 4. Then, the spectral radiance of light of each set wavelength is calculated as an analog amount in the measuring device body B,
The analog quantity related to the spectral radiance is converted into a digital signal by the A / D converter 5 and then input to the arithmetic unit D such as a personal computer. On the other hand, the distance (h) between the object M to be identified and the measuring head H by the ultrasonic sensor 2
Is detected, and the analog amount related to the measured distance (h) is also amplified by the amplifier 6, converted into a digital signal by the A / D converter 5 and input to the arithmetic unit D.

Next, the arithmetic unit D will be described. This arithmetic unit D includes a spectral radiance for each set wavelength at a reference measurement distance of a reference object such as a white surface,
A relational expression regarding the change of the spectral radiance with respect to the measurement distance of the reference object is input.

Here, a method of measuring the spectral reflectance of the object will be briefly described. First, a sample of the identified object,
Prepare a reference object such as a white surface that serves as a reference for measuring the spectral reflectance, irradiate this reference object with a standard light source from the reference measurement distance, and measure the spectral radiance at that time for each arbitrary wavelength. To do. Further, from the reference measurement distance, the reference object is moved by a constant distance in both + (plus) and-(minus) directions, and the spectral radiance at that distance is measured in the same manner as above. By repeating the above measurement, the relational expression between the measurement distance between the measuring head of the measuring instrument and the reference object and the spectral radiance is obtained for each arbitrary wavelength. Next, regarding the sample of the object to be identified, in the same manner as the above-mentioned reference object, the spectral radiance is measured for each arbitrary wavelength and by changing the distance between the sample and the measuring head as described above. To do. The distance between the reference head or the sample of the object to be identified and the measuring head can be easily measured by the ultrasonic sensor. Then, when the measured distance is the same, the spectral radiance of the sample of the object to be identified is compared with that of the reference object, and the spectral reflectance of the sample at that measured distance is calculated.

FIG. 3 is a diagram showing the measured waveforms of the spectral reflectance with respect to the wavelengths of orchardgrass (grass), Rumex japonicus (broadleaf weed), dry soil, and wet soil. It can be seen that the spectral reflectance of the plant body is different in magnitude at each wavelength, but the waveform changes are almost the same. The characteristics of the waveform are
There is a mountain near 550 nm and a valley near 670 to 680 nm, which rises sharply from 700 to 750 nm, then drops with a slight swell, and large valleys again exist near 1400 nm and 1900 nm. Regarding soil,
Compared with plants, there are almost no peaks or valleys, the spectral reflectance is low, and the waveform gradually rises to the right. It should be noted that the slope of moist soil having a high water content becomes more gentle. For this reason, the spectral reflectance of the object to be identified is measured at two locations, near the peak and valley of the spectral reflectance of the plant, around 550 nm and around 670 nm, and the magnitudes are compared to identify It is possible to determine whether the object is a plant or soil (if the spectral reflectance around 550 nm is higher than around 670 nm, the object to be identified is a plant, and vice versa). Becomes soil).

For this reason, the arithmetic unit D includes a plurality of different grasses (broad grasses and broad-leaved weeds to be distinguished) and soils within a wavelength range of 550 to 1300 nm (measurement heads shown in FIG. 1) as shown in FIG. In case of H, 6 types of set wavelength (G
The respective reference spectral reflectances in _{1 to} G ₆ ) are input. The reference spectral reflectance is calculated by the method described above by setting the distance between each measuring plant and the soil and the plants collected as samples as the reference measurement distance.

Further, in the above-described optical identification method, the measuring head of the measuring instrument and the object to be identified move relative to each other in a state where the measurement distance between the two changes. As described above, when the measurement distance between the measurement head and the object to be identified changes, the spectral radiance of the object to be identified also changes, so that the arithmetic unit D has a measurement head of the measuring instrument and a white surface. A relational expression regarding the change of the spectral radiance with respect to the change of the measurement distance from the reference object is input. This relational expression is easily calculated by using this data, because various data are obtained when measuring the spectral radiance of a reference object such as a white surface by changing the measurement distance by the above method. To be done. For example, as shown in FIG. 5, at each set wavelength (G _{1 to} G ₆ ), a relational expression [R ₁ = F ₁ (h) to about a change in spectral radiance before and after the reference measurement distance (h ₀ ) R ₆ = F ₆ (h)],
This is input to the arithmetic unit D in advance.

Then, as shown in FIG. 6, the measuring instrument main body B, the measuring head H and the arithmetic unit D are placed on the work vehicle 7 to run on the pasture, and the pasture (orchardgrass) having a relatively small leaf width. ) And a broad-leaved weed (Ezonoginagi). In addition, the work vehicle 7,
A squeezing device 8 for squeezing the plant immediately before the measurement area to bring it closer to a two-dimensional state (planar) is provided. Even when the plants (pastures and broad-leaved weeds) are squeezed by the squeezing device 8 to approach a two-dimensional state, some undulations remain and
Since the soil has some unevenness, the measurement distance between the grass and the broad-leaved weeds and the soil and the measuring head H of the measuring instrument changes, but the measuring head H includes the grass, the broad-leaved weeds, and the soil. Is equipped with an ultrasonic sensor 2 for measuring the distance, and the arithmetic unit D relates to a reference object such as a white surface, and a relational expression [R ₁ = F ₁ (h) to R ₆ = F
₆ (h)] is input in advance, and the respective spectral reflectances of the grass, the broadleaf weed, and the soil at each set wavelength (G _{1 to} G ₆ ) are
A spectral radiance at each set wavelength detected by each of the plurality of detectors C _{1 to} C ₆ constituting the measurement head H, and a measurement distance detected by the ultrasonic sensor 2.
It is calculated based on the relational expression. Then, the calculated spectral reflectance of the identified object is arithmetically processed by the following method,
It distinguishes grass from broadleaf weeds.

Here, an experimental example will be given on the relationship between the size of the measurement area of the measuring head H and the moving speed of the measuring head H when the measuring head H is moved to measure the spectral reflectance of the object to be identified. Explain. As shown in FIG. 7, a strip-shaped orchard glass OG with a width of 1 cm and a width of 5
Within the typical grassland, a typical grassland in which the soil S is provided is formed between the same strip-shaped Ezonoginagi X of cm and the measuring head H having the measuring area A formed of a circle of radius r is moved at the speed V. The respective spectral reflectances of Orchardgrass OG, Ezo Nogeki X, and Soil S were continuously measured. In order to show that the spectral reflectances can be distinguished even if they are the same, the spectral reflectances of Orchardgrass OG and Ezo Noshiki X are both approximately 50% (these are approximately equal to the spectral reflectances at wavelengths near 780 nm). The spectral reflectance of soil was set to 10%. Further, as shown in FIG. 8, when the measurement area A straddles both the Rumex japonicus X and the soil S, the spectral reflectance RS of the soil
Is 10%, and the spectral reflectance RX of Rumex japonicus is 50%,
The combined spectral reflectance is RR, the area of the measurement area is AA, the area of the soil is AS, and the area of Rumex japonicus is A when the measurement area spans both the Ezo nose grass X and the soil S.
X, the combined spectral reflectance RR is calculated by the formula [RR = RS ×
(AS / AA) + RX × (AX / AA)].
For example, when the measurement area has the same area across both the soil S and the Rumex japonicus X as illustrated, the combined spectral reflectance RR is calculated to be 30%.

Then, the spectral reflectance is continuously measured by changing the measuring area A of the measuring head H and the moving speed V of the measuring head H variously in the above-mentioned typical meadow.
An experiment was conducted as to whether or not Orchardgrass OG and Spiraea japonicus X, which have almost the same spectral reflectance, can be discriminated, and the results are shown in FIGS. 9 to 32. The lower threshold was assumed to be 45%. Here, FIGS. 9 to 16 show the spectrum of the object to be identified when the moving speed V of the measuring head H is fixed to 0.2 cm / s and the radius r of the measuring area is sequentially increased from 0.2 cm to 5.0 cm. It is a graph which shows the result of having measured reflectance continuously. In the combination of the measurement area radius r and the moving speed V,
For those that can and cannot distinguish between Orchardgrass and Rumex japonicus, circles and crosses are attached at the top of each graph. From this result, when the measurement area radius r is as small as 0.2 cm (see FIG. 9), both the orchardgrass and the Rumex japonicus are detected, and it is difficult to distinguish both plant bodies, and the measurement area radius r = In the range of 0.5 to 2.0 cm, the spectral reflectance of Rumex japonicus appears as a constant straight line, and only the spectral reflectance of orchardgrass is below the threshold, and both plants can be distinguished. When the radius r of the measurement area exceeds 3.0 cm, even the spectral reflectance of Rumex japonicus falls below the threshold value, and it has been found that it is difficult to distinguish both plants. The fact that the spectral reflectance of Ezo no Rumegi appears as a constant straight line means that the measured spectral reflectance was included a predetermined number of times within the range of the spectral reflectance that is considered to be Ezo no Rume. It can be seen from the above that there is Rumex japonicus in this portion.

17 to 24, the object to be identified in the case where the moving speed V of the measuring head H is fixed to 0.6 cm / s and the radius r of the measuring area is successively increased from 0.2 cm to 5.0 cm. 3 is a graph showing the results of continuous measurement of the spectral reflectance of. From this result, the measurement area radius r
= 0.2 to 1.5 cm, the spectral reflectance of the orchardgrass part is sharply mountain-shaped or is below the threshold, and the linear reflectance part of the Ezo-no-Tsumugi part is present. It is possible to identify the body,
From the part where the measurement area radius r exceeds 2.0 cm to the spectral reflectance of the part of Rumex japonicus, there is a dull mountain shape or it becomes less than the threshold value, and it was found that it is difficult to distinguish both plant bodies.

25 to 32, the object to be identified in the case where the moving speed V of the measuring head H is fixed to 1.0 cm / s and the radius r of the measuring area is gradually increased from 0.2 cm to 5.0 cm. 3 is a graph showing the results of continuous measurement of the spectral reflectance of. From this result, the measurement area radius r
In the range of 0.2 to 1.0 cm, the spectral reflectance of the orchardgrass part is below the threshold value, and the linear reflectance part of the Ezonotigaki part is present, so both plants can be distinguished. There is a measurement area radius r
It was found that it is difficult to discriminate between the two plant bodies, since the spectral reflectance of the Ezo no Rumegi from the part exceeding 2.0 cm becomes dull peak-shaped or below the threshold.

From the above measurement results, the following can be said. When the measurement area is smaller than a certain value, the measurement area is smaller than the width of the narrow orchard glass, and it is difficult to identify unless the moving speed is high. If the area is larger than a certain amount, not only the plants to be identified in the measurement area, but the soil before and after that is always included, the spectral reflectance of both plants is below the threshold value, and it becomes impossible to identify, It can be seen that it is indistinguishable regardless of the moving speed. As a result, when the moving speed is slow at 0.2 cm, the measurement area radius is 0.7
It is possible to distinguish Orchardgrass and Rumex japonicus at ~ 2.0 cm, and even if the moving speed is slightly faster than this, the tendency is almost the same, but the measurement area radius is 2.
When it exceeded 0 cm, the measured waveform in the part of Rumex japonicus became mountain-shaped, which made it difficult to distinguish between the two plants. In addition, when the moving speed is as fast as 1.0 cm / s, the smaller the measurement area area, the higher the possibility of identification, and when the measurement area radius reaches 1.5 cm, the identification becomes somewhat difficult. Taken together, it seems that a measurement area radius of 0.7 to 1.2 cm is appropriate for distinguishing leaves of plant bodies with widths of 1 cm and 5 cm, and the spectral reflectances of both plants are almost the same. Even if they are the same, it can be seen that if the size of the leaf is different, the combination of the measurement area of the measurement head H and the moving speed of the measurement head H can be discriminated from each other.

In the above example, since the two types of plants to be discriminated have substantially the same spectral reflectance, the threshold values are the same, but when the spectral reflectances are different,
It is necessary to set a threshold value corresponding to each spectral reflectance, and the spectral reflectance of legumes such as red clover around 780 nm (near infrared) may be higher than that of Ezo grasshopper when the leaves overlap. In this case, it is necessary to provide not only the lower threshold value but also the upper threshold value. In addition, what has been described above is the size of the measurement area and the speed of the grass, which is the object to be identified (Orchard grass).
It is a theoretical explanation that even if the spectral reflectance of both is similar, it is possible to distinguish them by selecting the optimum one corresponding to the leaf width of broadleaf weeds (Ezonoginugi). Alternatively, the leaves of broad-leaved weeds extend in any direction, and if the measurement area is moved in any direction corresponding to this, it is possible to identify only broad-leaved weeds corresponding to the actual meadow. .

In the above example, the spectral reflectances at one specific wavelength are continuously measured, and the two types of plants whose spectral reflectances are close to each other and whose leaf sizes are greatly different are identified. However, if the spectral reflectances at a plurality of different wavelengths are combined, highly accurate identification is possible. For example, 550nm (green), 670nm (red), 78
An identification method using three kinds of wavelengths of 0 nm (near infrared) will be described. In exactly the same manner as above, while moving the work vehicle 7, the measurement head H measures the spectral reflectance at each of the three types of wavelengths described above. First, as shown in the flow chart of FIG. 33, the spectral reflectances of the respective wavelengths in and are compared, and if is higher, then the plant area occupies a larger area than the soil in the measurement area. After making a determination, the process proceeds to the next calculation step. In the opposite case, it is determined that the soil or dead leaves occupy most, and the process returns to the original. After cutting grass, since there are many dead leaves at the root of the plant, it will be shaken off considerably by this operation. In the actual field test, grass was also detected at one wavelength, but it was not detected when measuring at three wavelengths.

If the spectral reflectance at the wavelength of is higher than that at the wavelength of, and it is determined that the proportion of plants in the measurement area is high, 1
The calculation proceeds in the same way as for the wavelength. It is judged whether the spectral reflectance at the wavelength is within the threshold value, and if it is within the threshold, it is "1" to indicate that it is edging.
Is output, and if it is not entered, "0" is output and the process returns to the original. When it returns to the original state, the same calculation is continuously performed, and if the spectral reflectance at the wavelength of is within the threshold value, the output “1” is added to be “2”. This operation is continuously repeated, and only when the number of times of determination reaches a predetermined value (for example, 5 times), it is finally determined that the object to be identified is Ezo no Rume, and a signal is output. On the other hand, when moving from Ezo no Gravel to other objects to be identified (grass grass, soil, etc.),
If the reverse operation is performed and the Ezo-no-Tsumugi is not detected by the predetermined value (for example, 5 times) continuously, the operation of stopping the signal is performed. FIG. 34 is a graph in the case where the spectral reflectances at the wavelengths of, and are continuously measured in the meadow, and the spectral reflectance at the wavelength of is larger than that at the wavelength of It can be seen that Rumex japonicus is detected in the part of the data number where the spectral reflectance at the wavelength is over 45%.

Therefore, the measurement area patterns are classified into the following three types. When the spectral reflectance at the wavelength of (1) is larger than that at the wavelength of and the spectral reflectance at that wavelength is within the threshold value, the inside of the measurement area is edgy.
If the spectral reflectance at the wavelength of (2) is larger than that at the wavelength of and the spectral reflectance at that wavelength is smaller than the lower side of the threshold or larger than the upper side of the threshold, within the measurement area. Are plants other than Rumex japonicus. When the spectral reflectance at the wavelength of (3) is smaller than that at the wavelength of, the measurement area is soil.

Whether or not a signal is output is classified into the following cases depending on whether or not the above three types of patterns appear in succession. (1) In the case of Ezo-no-Tsumugi, which exceeds the preset value continuously until the last time, and again this time, the signal is continuously output. (2) Up to the last time, the signal exceeds the default value, and in the case of other plants this time, the signal is continuously output.
(3) Up to the last time, the signal is continuously output when the value exceeds the preset value, and this time it is soil. (4) No signal is output in the case of soil or other plants up to the last time, and this time in the case of Rumex japonicum.
(5) No signal is output in the case of soil or other plants continuously until the last time, and again in the case of soil or other plants.
(6) If it is less than the default value and it exceeds the default value for the last time,
A signal is output.

[0026]

INDUSTRIAL APPLICABILITY The present invention provides an optimum measurement area of a spectral reflectance measuring instrument based on the average width of each leaf of broadleaf weeds and pastures to be distinguished, and an optimum measuring area If the moving speed is set in advance and a plurality of objects to be identified are included in the moving measurement area, the composite spectral reflectance is calculated and the spectrum of the objects to be identified on the meadow at a specific wavelength is calculated. The reflectance is continuously measured, and the measured spectral reflectance is included only within a predetermined number of consecutive times within the range of the spectral reflectance that is regarded as a preset broadleaf weed. This is a method of distinguishing an identification object as a broad-leaved weed, so if there is a certain difference in leaf width between the grass and the broad-leaved weed, they will be identified even if their spectral reflectances are close to each other. It becomes possible. In this case, if the spectral reflectances are respectively measured at a plurality of different wavelengths, the identification accuracy will be further enhanced.

[Brief description of drawings]

FIG. 1 is a principle diagram of an optical identification method of an object using a spectral reflectance measuring device.

2 is a bottom view of the measuring head H. FIG.

FIG. 3 is a diagram showing measurement waveforms of spectral reflectance with respect to wavelength for orchardgrass (grass), Rumex japonicus (broadleaf weed), dry soil, and wet soil.

FIG. 4 is a diagram showing reference spectral reflectances with respect to a plurality of specific set wavelengths (G _{1 to} G ₆ ) of orchard grass, Rumex japonicum, dry soil and wet soil, which are input to the arithmetic device D.

FIG. 5 is a diagram showing a change in spectral radiance with respect to a change in measurement distance at a specific set wavelength.

FIG. 6 is a schematic diagram showing a state in which the work vehicle is traveling to continuously measure the spectral reflectance of the meadow.

FIG. 7 is a plan view of a schematic meadow.

FIG. 8 is a diagram showing a principle of calculating a composite spectral reflectance.

FIG. 9: Measuring area radius r = 0.2 cm, moving speed V = 0.2c
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow in m / s.

FIG. 10: Measurement area radius r = 0.5 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 11: Measurement area radius r = 0.75 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 12: Measurement area radius r = 1.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 13: Measurement area radius r = 1.5 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 14: Measurement area radius r = 2.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 15: Measurement area radius r = 3.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 16: Measurement area radius r = 5.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 2 cm / s.

FIG. 17: Measurement area radius r = 0.2 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 18: Measurement area radius r = 0.5 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 19: Measurement area radius r = 0.75 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 20: Measurement area radius r = 1.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 21: Measurement area radius r = 1.5 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 22: Measurement area radius r = 2.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 23: Measurement area radius r = 3.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 24: Measurement area radius r = 5.0 cm, moving speed V = 0.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 6 cm / s.

FIG. 25: Measurement area radius r = 0.2 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 26: Measurement area radius r = 0.5 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 27: Measurement area radius r = 0.75 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 28: Measurement area radius r = 1.0 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 29: Measurement area radius r = 1.5 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 30: Measurement area radius r = 2.0 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

[FIG. 31] Measuring area radius r = 3.0 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 32: Measurement area radius r = 5.0 cm, moving speed V = 1.
It is a figure which shows the change of the spectral reflectance with respect to the elapsed time of a typical meadow at 0 cm / s.

FIG. 33 is a flowchart for discriminating between orchard grass (grass) and Ezo regula (wideleaf weeds) using three kinds of wavelengths.

FIG. 34 is a diagram in which the spectral reflectance of a meadow is continuously measured using three types of wavelengths.

[Explanation of symbols]

 A: Measuring area B: Measuring device D: Computing device H: Measuring head OG: Orchardgrass (grass) S: Soil V: Moving speed of measuring head X: Rumex japonicus (broad-leaved weed) r: Radius of measuring area 7: Work car

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Masahiro Kamei 1-40, Nisshin-cho, Omiya-shi, Saitama 2 Within the Institute for the Promotion of Specified Industrial Technology of Biological Systems (72) Manabu Soge 2828 Natsumi, Nabari-shi, Mie Co., Ltd. In Takakita (72) Inventor Katsumi Yamazaki 2828 Natsumi, Nabari City, Mie Prefecture Takakita Incorporated (72) Inventor Tsuyoshi Okura 5-15-14 Ogikubo, Suginami-ku, Tokyo Within Opt Research Inc.

Claims (2)

[Claims] 1. A grass having a relatively small leaf width and a leaf having a leaf width smaller than that of the grass are measured by continuously measuring the spectral reflectance of the object to be identified while moving the spectral reflectance measuring device along the ground surface of the pasture. A method for distinguishing broad-leaved weeds that are much wider in width, which is an optimal measurement area of the spectral reflectance measuring device based on the average width of each leaf of the broad-leaved weeds and grass to be identified, and the measurement. An optimum moving speed of the measuring device for an area is determined in advance, and when a plurality of objects to be identified are included in the moving measuring area, the composite spectral reflectance thereof is calculated, and The spectral reflectance of the identified object of the pasture at the wavelength is continuously measured, and the measured spectral reflectance is continuously set within the range of the spectral reflectance that is regarded as a preset broadleaf weed. Only when the object is identified as a broadleaf weed Identifying broadleaf weeds spectral reflectance pasture in utilizing, characterized by another. 2. A grass having a relatively small leaf width and a leaf having a leaf width smaller than that of the grass are measured by continuously measuring the spectral reflectance of the object to be identified while moving the spectral reflectance measuring device along the ground surface of the pasture. A method for distinguishing broad-leaved weeds that are much wider in width, which is an optimal measurement area of the spectral reflectance measuring device based on the average width of each leaf of the broad-leaved weeds and grass to be identified, and the measurement. The optimum moving speed of the measuring device for the area is determined in advance, and three or more kinds of wavelengths capable of distinguishing grass, broad-leaved weeds and soil are selected, and a plurality of wavelengths are set in the moving measuring area. If the identified object is included, the combined spectral reflectance thereof is calculated, and the spectral reflectance of the identified object of the meadow at each wavelength is continuously measured to obtain two specific wavelengths. By comparing the spectral reflectance in the If it is determined that it is not soil, then the spectral reflectance of the identified object at all or part of the remaining wavelength was considered to be that of broadleaf weeds for the specified number of consecutive times. Only in some cases, the object to be identified is identified as a broadleaf weed, and a method for identifying a broadleaf weed in a meadow using spectral reflectance.