JP2815760B2 - Crop row detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image pickup means for picking up an image of a predetermined range of field scenes including crops arranged in a plurality of rows, and an area for extracting a region corresponding to the crop based on image pickup information by the image pickup means. A crop provided with an extraction unit and a calculation unit for obtaining an unplanted side region corresponding to a crop arranged in a row adjacent to the unplanted region in the field based on the region information extracted by the region extraction unit; The present invention relates to a column detection device.

[0002]

2. Description of the Related Art A crop row detecting apparatus of this kind, when planting seedlings as crops at a set interval in a unit of a plant, such as a rice planting machine, is arranged in the advancing direction of the machine. In order to obtain control information for automatic running along the seedling row, of the areas extracted corresponding to the seedlings arranged in a plurality of rows on the imaging screen, a row is formed adjacent to the unplanted area of the field. The area corresponding to the lined seedlings is determined. And, based on the information of the area adjacent to the unplanted area, for example, to detect the position and direction of the already planted seedling row with respect to the aircraft position by approximating the line connecting those areas to a straight line or a curve, or The control information is obtained by detecting the position of the planted seedling row with respect to the machine position from the information of the average position (for example, with respect to the center of the screen) of those areas in the screen.

Conventionally, when an area corresponding to a seedling arranged in a row adjacent to an unplanted area in the above-mentioned field is determined, for example, the area closest to the unplanted area on each horizontal scanning line on the horizontal axis of the screen is determined. It is determined whether or not the area is located at a position adjacent to the unplanted area based on the determination (for example, see Japanese Patent Application No. 3-95120).

[0004]

However, in the above-mentioned prior art, all the area information originally existing on the imaging screen is mutually arranged in a predetermined (determined by the configuration such as the planting interval of the seedlings and the magnification of the imaging means). Despite having the relationship, the area other than the area closest to the unplanted area is simply discarded as a result of the position determination on the horizontal scanning line, and the arrangement information of each of the above areas is discarded. Was not being used effectively. Therefore, it is possible that proper detection accuracy may not always be ensured by simply finding the region located closest to the unplanted region on each horizontal scanning line on the screen as a region adjacent to the unplanted region. In addition to the position determination on the line, for example, it is necessary to perform processing for securing accuracy by removing, for example, a region where the position of a vertically adjacent region determined to be located closest to the unplanted region is significantly shifted. There was a problem that there was.

[0005] The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to obtain all areas on an imaging screen when obtaining an area corresponding to a seedling arranged in a row adjacent to an unplanted area. An object of the present invention is to improve the accuracy in obtaining the area by effectively utilizing the area information.

[0006]

According to a first characteristic configuration of a crop row detecting apparatus according to the present invention, the arithmetic means extends along the row direction of the crop on the image screen of the imaging means, and Having a predetermined width in a direction intersecting with the column direction of the crop, and juxtaposed so as to correspond to an arrangement of a previously predicted row of the crop when the imaging unit captures the crop in a setting appropriate state. A plurality of elongate masks, and the plurality of elongate masks, at each of a plurality of moving points separated by a set distance in a direction intersecting with the row direction of the crop, at a specific point in the row direction Is moved to a plurality of mask movement positions specified as positions separated by a predetermined angle from the center, and at each of the plurality of mask movement positions, the number of the regions located inside each of the plurality of long masks is calculated. Count and duplicate The count value of the area is configured to determine the non-plant-side territory <br/> area based on the position information of the plurality of elongated mask in the mask moved position with the maximum among the mask moving position It is in the point.

A second characteristic configuration is that a crop presence / absence detection mask for detecting the presence / absence of the crop is located at a position closer to the unplanted area than the plurality of elongated masks. Provided so as to have an arrangement relationship corresponding to the arrangement of the previously predicted columns of the crops with respect to the shape mask, wherein the calculating means, when the region is present inside the crop presence / absence detection mask, The present invention is configured such that a predetermined value is subtracted from a count value of the region located inside each of the plurality of long masks.

A third characteristic configuration is that the imaging means is provided so as to image the field scene diagonally forward and downward, and the plurality of long masks are arranged in a plurality of rows of the crop. The point is that a line extending in the forward direction in the traveling direction is arranged so as to extend radially from an intersection where the line intersects at one point.

In a fourth characteristic configuration, the imaging means is provided so as to image the field scene vertically downward, and the plurality of long masks are arranged so as to be parallel to each other. It is in the point.

[0010]

According to the first characteristic configuration of the present invention, when a crop is taken in an appropriate setting state, the crop extends in the crop row direction and has a predetermined width in a direction intersecting the crop row direction. A plurality of elongate masks juxtaposed so as to correspond to the arrangement of the pre-predicted rows of crops are arranged in a row direction at each of a plurality of moving points separated by a set distance in a direction intersecting with the row direction of the crops. Is moved to a plurality of mask movement positions specified as a rotated position separated from the specific point by a predetermined angle, that is, a rotated position, and is located inside each of the plurality of long masks at each of the plurality of mask movement positions. The number of regions is counted. Then, based on the position information of the plurality of long masks at the mask movement position where the count value of the region is the largest among the plurality of mask movement positions, for example, The region located inside the long mask located adjacent to the unplanted region side as an unplanted side region corresponding to crops arranged in a row adjacent to the unplanted region in the field Ask.

That is, as shown in FIG. 4, a region Ta located inside each of a plurality of long masks M1 and M2.
At the mask movement position where the number of n and Ta1 is maximum, a plurality of long masks M1 and M2 arranged so as to properly correspond to the arrangement of the crop rows and the crop actually imaged in the screen are displayed. Since the row and the row are in the best matching state, the position information of the plurality of long masks M1 and M2 at the mask movement position best represents the position information of the crop row at that time. Therefore, based on the position information of the plurality of long masks M1 and M2 at the mask movement position , for example, the most
Long mask M located adjacent to the unplanted area side of the field
If the region Tan located inside 1 is determined as a region corresponding to crops arranged in a row adjacent to the husband planting region N in the field, the detection accuracy can be increased.

According to a second characteristic configuration, FIG.
As described above, the plurality of long masks M1 and M2 and the crop presence detection mask M3 positioned on the unplanted area N side from the plurality of long masks M3 in the above-described setting appropriate state are similar to the above. When there is an area inside the crop presence detection mask M3 at each of the plurality of mask movement positions, the plurality of long masks M1 and M2 are respectively moved inside the crop existence detection mask M3. A predetermined value (for example, the above-described crop presence detection mask M3)
Is subtracted, and a mask moving position at which the count value of the region is the maximum among a plurality of mask moving positions is obtained. That is, the crop presence detection mask M3 is a mask for confirming that no region exists inside the crop presence mask. The mask movement position where the number of the regions is maximum is a plurality of long masks M1 and M2. Within the region T as much as possible
an, Ta1 are present and the crop presence detection mask M
3 shows the most matched state when there is no region as much as possible. Therefore, in this case, the region T for crops arranged in a row adjacent to the unplanted region N in the field
By using the arrangement relationship that no area exists on the unplanted side than an, the detection accuracy when obtaining the area corresponding to the crops arranged in a row adjacent to the unplanted area N in the field can be increased. It is.

According to the third characteristic configuration, when the field scene is imaged diagonally forward and downward, the image is displayed as if the lines extending in the forward direction in the traveling direction of each crop radiated from the intersection where one line intersects at one point. A plurality of long masks, or a plurality of long masks and the crop presence / absence detection mask are also arranged in the same manner in accordance with the image taken on the top. And also in this case, at the mask movement position where the number of the areas is the maximum, a plurality of long masks or a plurality of long masks arranged and arranged so as to appropriately correspond to the arrangement of the crop rows as described above. Since the arrangement of the long mask and the crop presence / absence detection mask and the crop row in the screen are in the most matched state, the field of the field is determined based on the position information of the plurality of long masks at the mask movement position. If a region corresponding to crops arranged in a row is obtained adjacent to the unplanted region, the detection accuracy can be increased even when the field scene is imaged diagonally forward and downward.

According to the fourth characteristic configuration, when the field scene is imaged vertically downward, the crop is imaged on the screen in a state where the row directions of the crop are arranged in parallel.
A plurality of long masks, or a plurality of long masks and the crop presence / absence detection mask are similarly arranged. And also in this case, at the mask movement position where the number of the regions is the maximum, a plurality of long masks arranged and configured to appropriately correspond to the arrangement of the crop rows as described above,
Alternatively, since the arrangement of the plurality of long masks and the crop presence / absence detection mask and the crop row in the screen are most matched, the position information of the plurality of long masks at the mask moving position is included. If the area corresponding to the crops arranged in a row adjacent to the unplanted area of the field is determined based on the field, the detection accuracy can be increased even when the field scene is imaged vertically downward. .

[0015]

Therefore, according to the first characteristic configuration of the present invention, the arrangement relationship between a plurality of crop rows on the screen is effectively utilized, and the crop rows are arranged adjacent to the unplanted area of the field. The detection accuracy of the region corresponding to the crop can be improved, and thus, it can be favorably used as steering control information at the time of automatic traveling of the work vehicle or the like.

Further, according to the second characteristic configuration, not only the arrangement relationship between the plurality of crop rows on the screen but also the position closer to the unplanted side than the crops arranged in a row adjacent to the unplanted area of the field. Makes effective use of the layout relationship that no crop exists,
The detection accuracy of the region corresponding to the crops arranged in a row adjacent to the unplanted region in the field is further improved, and the effect of the first characteristic configuration can be further enhanced.

According to the third characteristic configuration, even when a field scene is imaged diagonally forward and downward, the direction of the crop row on the screen is appropriately adjusted to be radial, and The region corresponding to the crops arranged in a row adjacent to the unplanted region is accurately detected, and the effect of the first or second feature configuration can be reliably realized.

Further, according to the fourth characteristic configuration, even when the field scene is imaged vertically downward, it is possible to properly cope with the arrangement of the crop rows on the screen in a parallel state. The area corresponding to the crops arranged in a row adjacent to the unplanted area is accurately detected, and the effect of the first or second characteristic configuration can be reliably realized.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the present invention is applied to a seedling row detecting device for detecting the position of a row of seedlings planted in a field by a rice planting machine will be described below with reference to the drawings.

As shown in FIGS. 7 and 8, a seedling planting device 2 is provided at the rear of an airframe V having both front wheels 1F and rear wheels 1R operable for steering operation. A plurality of already planted seedlings Tn, T1, T2 as crops arranged in a plurality of rows are planted in the device 2. Then, the already planted seedlings Tn, T
A color-type image sensor S1 as an image pickup means for picking up an image of a field scene in a predetermined range including the areas T1 and T2 is provided on the front side of the machine body V.

The mounting structure of the image sensor S1 will be described. The image sensor S1 is mounted on the front end of the support member 4 which projects laterally at the front side of the body V, and a plurality of the image sensors S1 are mounted on the lateral outside of the body V. Planted seedlings T lined up in a row
The row of n, T1, and T2 is provided so as to image diagonally forward and downward. That is, in a state where the body V is properly aligned with the rows of the plurality of planted seedlings Tn, T1, T2 arranged in the body traveling direction, the body V is arranged in a row adjacent to the unplanted area N of the field. The line segment L corresponding to the existing planted seedlings Tn is aligned with a running reference line La passing through the center of the field of view of the image sensor S1 in the front-rear direction.

A plurality of work steps from one end of the field to the other end are set in a state of being arranged in parallel in the machine width direction. In each work step, based on the image information of the image sensor S1, The steering control is performed so as to automatically travel along the already planted seedling row. However, as the end of one work stroke is reached, the vehicle is automatically turned in a state of turning 180 degrees toward the start end of the next work stroke adjacent to the work stroke.

Accordingly, the running direction of the machine V with respect to the field is reversed every time the machine travels one stroke, and the positions of the planted seedlings Tn, T1, and T2 with respect to the machine V are in a state of being horizontally reversed. The image sensor S1 is provided with an airframe V
One is provided for each of the left and right, and the sensor to be used is switched left and right for each stroke. In FIG. 7,
Since the positions of the already planted seedlings Tn, T1, and T2 are on the right side of the body V, the image sensor S1 on the right side is used.

To explain the structure of the body V, as shown in FIG. 1, the output of the engine E is transmitted to each of the front wheel 1F and the rear wheel 1R via a transmission 5, and as the operation state of the speed change operation state corresponds to a preset speed, the shifting state detection potentiometer R ₃ provided, and, on the basis of the detection information of the shifting state detecting potentiometer R _3, shift It is configured to drive the electric motor 6 for use.

Further, the front wheel 1F and the rear wheel 1R are configured to be individually power-steered by hydraulic cylinders 7F and 7R, respectively, and the steering angle detecting potentiometers R ₁ and R interlocked with the steering operation of the wheels. _The electromagnetically operated control valves 8F and 8R for operating the hydraulic cylinders 7F and 7R are configured to be driven such that the steering angle detected by ₂ becomes the target steering angle.

Accordingly, a parallel steering system in which the front wheel 1F and the rear wheel 1R are steered in the same phase and at the same angle,
Three types of steering types, that is, a four-wheel steering type in which the front wheel 1F and the rear wheel 1R are steered in opposite phases and at the same angle, and a two-wheel steering type in which only the front wheel 1F is turned, can be selected and used. Has become.

However, when the steering operation is automatically performed based on the imaging information of the image sensor S1, the two-wheel steering system is used, and when one work process is completed and the next work process is performed, The four-wheel steering system and the parallel steering system are used. In FIG 1, S ₂ is the distance sensor for detecting a travel distance based on the output rotational speed of the transmission 5.

Next, based on the imaging information of the image sensor S1, the regions Tan, Ta1, Ta2 corresponding to the planted seedlings Tn, T1, T2 are extracted, and the regions Tan, Ta1, Ta2 of the extracted regions Tan, Ta1, Ta2 are extracted. Based on the information, the planted seedlings T arranged in a row adjacent to the unplanted area N in the field
In order to obtain the area Tan corresponding to n, and to approximate the line L connecting the obtained areas Tan, that is, the line L corresponding to the already planted seedlings Tn arranged in a row adjacent to the unplanted area N in the field. Will be described.

As shown in FIG. 1, the image sensor S
Numeral 1 is configured to output the three primary color information R, G, B separately, and from the green information G including the color components of the seedlings Tn, T1, T2 to the blue color not including the color components of the seedlings Tn, T1, T2. By subtracting the information B and binarizing, the seedling T
Areas Tan, Ta1, Ta2 corresponding to n, T1, T2
Is configured to be extracted.

In addition, a subtractor 9 for subtracting the blue information B from the green information G in the form of an analog signal,
The output of the subtracter 9 is binarized based on a preset threshold value corresponding to the colors of the seedlings Tn, T1, and T2, and binarized information corresponding to the regions Tan, Ta1, and Ta2 is output. The comparator 10 outputs the output signal of the comparator 10 at a preset pixel density (32 × 32 pixels / 1
The image memory 11 stores the image information corresponding to the seedlings Tn, T1, and T2 stored in the image memory 11. An area Tan corresponding to the seedling Tn adjacent to the unplanted area N in the field is obtained, and information that approximates a line L connecting these obtained areas Tan to a straight line or a curve is obtained, and traveling control is performed based on the information. Each of the control devices 12 using a microcomputer is provided.

That is, the subtractor 9 and the comparator 10 correspond to the area extracting means 100 for extracting the areas Tan, Ta1, Ta2 corresponding to the seedlings Tn, T1, T2 as crops. Then, using the control device 12, based on the information of the regions Tan, Ta1, and Ta2 extracted by the region extracting means 100,
Not planted corresponding to seedlings Tn adjacent to the non-planted area N of the field
The calculation means 101 for obtaining the side area Tan is configured.

The calculating means 101 displays the seedlings Tn, Tn on the image screen of the image sensor S1.
The seedlings Tn and T2 extend along the row direction
1, a line segment L corresponding to the seedling Tn adjacent to the unplanted area N in the field, having a predetermined width in a direction intersecting with the column direction of the image sensor S1, The seedlings Tn, Tn are aligned with the running reference line La passing through the center of the field of view of S1 in the front-rear direction.
1 and T2, a plurality of long masks M1 and M2 juxtaposed so as to correspond to the arrangement of the previously predicted rows of the seedlings Tn, T1 and T2 are set. At each of a plurality of moving points separated by a set distance in a direction intersecting the row direction of the seedlings Tn, T1 and T2, the long masks M1 and M2 are each set at a predetermined angle around a specific point in the row direction. The mask is moved to a plurality of mask movement positions set as distant positions, and at each of the plurality of mask movement positions, the region Tan, Ta1, Ta2 located inside the plurality of long masks M1, M2, respectively. And counting the area Ta among the plurality of mask movement positions.
The region Tan (corresponding to the seedling Tn adjacent to the unplanted region N in the field) based on the position information of the plurality of long masks M1 and M2 at the mask movement position where the count value of n, Ta1, and Ta2 is the maximum. Is determined.

Next, based on the flowchart shown in FIG. 2, the configuration of each part will be described in detail while explaining the operation of the control device 12.

Each time the body V travels a set distance or every set time, an image pickup process is executed by the image sensor S1, and as shown in FIGS. Regions Tan, Ta1, and Ta2 corresponding to the seedlings Tn, T1, and T2 are extracted.

The extraction of the regions Tan, Ta1, and Ta2 will be further explained. Since the colors of the seedlings Tn, T1, and T2 are green, attention is paid to green information G among three primary color information obtained by imaging a field scene. Then, the value corresponding to the region where the seedlings Tn, T1 and T2 are imaged is larger than the value of the region where another portion such as the mud surface of the field is imaged. However, since natural light is almost totally reflected on the water surface, the reflected light contains all color components. Therefore, the value of the green information G corresponding to the area where the water surface is imaged is the seedling Tn, T1, T2.
Is larger than the value of the region where the other part is imaged, similarly to the region where the image is captured. However, since the seedlings Tn, T1, and T2 are green and the mud surface is brown or gray, the value of the blue component contained in the color component is low.

That is, focusing on the blue information B among the three primary color information obtained by imaging the field scene, the reflected light from the water surface includes all color components, so that the value of the area where the water surface is imaged is large. The values of the areas where Tn, T1, T2 and the mud surface are imaged become small. Therefore, when the blue information B is subtracted from the green information G, only the value of the area corresponding to the seedlings Tn, T1, and T2 becomes larger than the value of the area where the other part is imaged. , The regions Tan, Ta corresponding only to the seedlings Tn, T1, T2.
1, Ta2 information can be extracted.

In the figure, the seedlings Tn adjacent to the unplanted area N and the two rows of seedlings T sequentially arranged from the seedlings Tn toward the planted side are shown.
1 shows a case where three rows of seedlings Tn, T1, and T2 of T2 are imaged in the screen. In addition, each said seedling Tn,
The imaging direction and the screen size are set so that about four regions (seedlings) are present in each column direction of T1 and T2, and each region (seedling) is a unit of the pixel density (32 × 32 pixels / 1). Although the screen is composed of a plurality of continuous pixel groups, when counting the number of regions described later, each pixel is counted as one. Therefore, a pixel position located at the center of gravity of each region Tan, Ta1, Ta2 is determined as a representative point P of the region. As described above, the image sensor S1 is connected to the seedlings Tn and Tn.
Since a plurality of rows of the seedlings Tn, T1, and T2 are configured to be imaged obliquely forward and downward, the lines extending from the plurality of rows of the seedlings Tn, T1, and T2 to the front in the traveling direction intersect at one intersection VP. .

Accordingly, the plurality of long masks M1,
M2 is a mask M1 extending radially from the intersection VP along the seedling Tn adjacent to the unplanted area N in the imaging field of view in the setting appropriate state, and a seedling T1 located next to the already-planted side.
And a pair of masks M1 and M2, each of which includes two masks of a mask M2 extending radially along. The angle θk formed by the center line in the longitudinal direction of each mask corresponds to the arrangement of the previously predicted rows of the seedlings Tn, T1, and T2 when the seedlings Tn, T1, and T2 are imaged in the setting appropriate state. The width .DELTA..theta. Of the mask in the direction intersecting the row direction of the seedlings Tn, T1, and T2 is set to a width that sufficiently covers the width of each of the regions (seedlings) in the screen horizontal direction (for example, 180 degrees to n so that
It is set as an equally divided angle width.

Next, the plurality of mask movement positions of the pair masks M1 and M2 will be described. As shown in FIG. 4, the intersection point VP is set to a point on the left (ρ = 0) on an axis extending in the horizontal direction of the screen. ) And the right side point (ρ = XV) are moved to a plurality of moving points separated by a set distance (XV / m) obtained by dividing the paired masks M1 and M2 by m. At the intersection V
A predetermined angle Δθ equal to the mask width Δθ (18)
0 ° / n). Thus, the masks do not overlap at adjacent rotational positions.
It should be noted that the angle is defined as an angle θ (a counterclockwise direction is defined as a plus direction with respect to a longitudinal center line of the left mask M1 of the paired masks M1 and M2) with respect to a direction facing downward in the screen vertical direction. Yes). Therefore,
The intersection point VP corresponds to a specific point in the row direction of the seedlings Tn, T1, T2, and a plurality of mask movement positions are determined by the positions of the plurality of movement points of the intersection point VP and the angular positions rotated by the predetermined angle. Will be set. The left point (ρ = 0) is located a predetermined distance left from the left edge of the screen.
Further, the right point (ρ = XV) is set at a position on the right side of the screen at a predetermined distance from the right end, and the deviation of the steering direction of the body V from the setting proper state is large, so that the intersection point VP is within the screen width. We can deal with cases that don't fit in.

Next, a description will be given of a process of counting the number of the regions Tan, Ta1, Ta2 located inside the paired masks M1, M2 at each of the set mask moving positions. Here, the intersection point VP is located at the point on the left side (ρ = 0) on the axis along the horizontal direction of the screen, and a position where the rotation angle θ is −90 degrees is a starting point. Of the regions Tan, Ta1, and Ta2 inside the masks M1 and M2 (specifically, the representative points P are counted), and the counted value (eight in the example of FIG. 4) is (ρ, θ ) Store in a predetermined area of the table (see FIG. 5). Then, the rotation angle θ is set to the predetermined angle △ θ.
Similarly, the number of areas inside the masks M1 and M2 is counted at the position increased by (180 ° / n) , and the count value is stored in a predetermined area of the (ρ, θ) table. Is repeated until + exceeds 90 degrees. Rotation angle θ
Exceeds +90 degrees, the position of the intersection point VP is moved to the right of the set distance (XV / m) from the left point (ρ = 0), and the position where the rotation angle θ is set to −90 degrees at that position. At each angular position increased by the predetermined angle △ θ, the process of counting the number of areas inside the masks M1 and M2 and the process of storing them in the (ρ, θ) table are performed by setting the rotation angle θ to +90 degrees. Repeat until it exceeds. And the intersection VP
Is the point on the right side on the axis along the horizontal direction of the screen (ρ = XV)
Execute the process up to the time when it is located at.

When the process of counting the number of regions at each mask movement position and the process of storing the data in the (ρ, θ) table are completed, as shown in FIG. Based on the positional information of the paired masks M1 and M2 at the mask movement position corresponding to the area (ρM, θM), the left mask M of the paired masks M1 and M2
1 is determined as an area corresponding to the seedlings Tn arranged in a row adjacent to the unplanted area N in the field.
In order to obtain control information for steering control from the obtained information of the area Tan, here, a straight line Lx corresponding to the longitudinal center line of the left mask M1 of the pair masks M1 and M2 is used. To the image sensor S1
Is obtained as line segment information connecting the pixels Tan corresponding to the already planted seedlings Tn arranged in a row adjacent to the unplanted area N on the imaging surface.

Next, as shown in FIG. 6, a straight line Lx on the imaging surface is obtained by storing information on the shape and size of the imaging visual field A of the image sensor S1 on the ground surface measured in advance.
Positions a, on the imaging plane through which the straight line Lx passes on the imaging plane
Based on b and c, the information is converted into information of a straight line L on the ground surface. That is, the information is converted into information of a straight line L on the ground surface set as a value of the inclination ψ with respect to the traveling reference line La passing through the center in the width direction of the imaging visual field A in the front-rear direction and the position δ in the width direction. Become.

In other words, the front and rear positions (y = 0 and y = 3) of the imaging visual field A in a direction intersecting a straight line L corresponding to a line segment connecting the planted seedlings Tn adjacent to the unplanted area N.
The lengths l ₀ and l ₃₂ of the two sides in 2), the center of the screen (x = 16,
The length l ₁₆ of the imaging field of view A in the width direction at the position where y = 16) and the distance h between the front and rear sides are measured in advance and stored in the control device 12. become.

Then, a straight line Lx on the imaging surface is
X at positions a and b (positions where y = 0, y = 32) intersecting the x-axis corresponding to the two sides at the front and rear positions of the imaging field of view A
A coordinate value X ₀ , X ₃₂ and a position c (1) at which the straight line Lx intersects the x-axis passing through the center of the screen , and a value X _{16 of the} x coordinate ,
It is obtained from the following equation (ii). In addition, this (i)
i) in equation (1), θ represents the straight line Lx as the origin at the center of the screen.
Parameters in polar coordinates (see FIG. 13).
See). Xi− (ρ−Yi · sin θ) / cos θ (ii) Here, 0, 16, and 32 are substituted for Y1.

Then, based on the coordinate values on the x-axis obtained by the above equation (ii), the following formulas (iii) and (iv) are used to calculate the position δ in the lateral width direction with respect to the travel reference line La.
And the slope ψ, and the obtained values of the position δ and the slope ψ
The position information is calculated as the position information of the straight line L corresponding to the planted seedling Tn on the ground surface.

[0046]

(Equation 1)

Therefore, in the steering control for automatically moving the body V along the row of planted seedlings arranged in the body traveling direction, the inclination of the straight line L with respect to the running reference line La
And the position δ in the width direction are both close to zero.
The steering operation will be performed in the form of wheel steering.

The steering control will be described from the values of the inclination の of the straight line L with respect to the traveling reference line La and the position δ in the lateral width direction and the value of the current steering angle φ of the front wheel 1F. , based on (v) below formula, and sets the target steering angle θf of the front wheel 1F, and the current steering angle φ is detected by the steering angle detecting potentiometer R ₁ for the front wheel, the target steering The control valve 8F of the front wheel hydraulic cylinder 7F is driven so as to be maintained within the set dead zone with respect to the direction angle θf. θf = K ₁ δ + K ₂ ψ + K ₃ φ (v) K ₁ , K ₂ , and K ₃ are constants set in advance according to the steering characteristics.

[0049] To describe reaches the end of the working stroke for turning control for turning toward the starting end of the next working stroke, the travel distance detected by said distance sensor S ₂ is one of the working stroke As the distance exceeds the set distance set in accordance with the length of the ridge, although not described in detail, the position of the ridge located on the front side in the traveling direction is determined from the image information of the imaging means S1. To reach the end of the work process. When it is determined that the end of the work process has been reached, the planting operation by the seedling planting apparatus 2 is interrupted, the mode is switched from the two-wheel steering mode to the four-wheel steering mode, and during the set time,
By maintaining the maximum cutting angle, 1
Turn by 80 degrees, then switch to the parallel steering mode and maintain the maximum turning angle for the set time to correct the position in the fuselage lateral direction for the next work stroke and end the turn Will do. After the turn, return to the two-wheel steering mode,
The steering control in the next work process will be restarted.

Next, another embodiment of the structure of the long mask will be described with reference to FIG. In this example, the seedlings Tn, T1, T2, and Crop as crops are provided in addition to the paired masks M1 and M2 as the plurality of long masks.
A crop presence / absence detection mask M3 for detecting the presence / absence of T3 is located at a position closer to the unplanted area N than the pair masks M1 and M2.
The seedlings Tn, T1, with respect to the pair masks M1, M2.
The arrangement relationship corresponding to the arrangement of the columns T2 and T3 predicted in advance, that is, the center line in the longitudinal direction of the mask M1 adjacent to the unplanted area N of the paired masks M1 and M2 and the presence or absence of the crop are detected. The angle θk ′ between the mask M3 and the center line in the longitudinal direction is provided so as to satisfy the above-described arrangement relationship. The mask width of the crop presence / absence detection mask M3 is set to the same angular width Δθ as the pair masks M1 and M2.

In this case, the arithmetic means 101 sets the seedling Tn inside the crop presence detection mask M3 at each of the mask movement positions specified by the movement position of the intersection point VP and the rotation angle θ. , T1, T2, T
In the case where the regions Tan, Ta1, Ta2, and Ta3 corresponding to No. 3 exist, a value obtained by subtracting a predetermined value from the count value of the regions Tan, Ta1, Ta2, and Ta3 located inside the pair masks M1 and M2, respectively. The count value at the mask movement position is used as the count value, and the count value is defined as (ρ,
θ) It is configured to store in a predetermined area on the table. As the predetermined value, a value of the number of areas located inside the crop presence detection mask M3 is used, or a value obtained by multiplying the value of the number of areas by an appropriate weighting coefficient is used. The flowchart of the control operation in this case is the same as that shown in FIG. 2 except that the count value for the area is obtained through the above-described subtraction processing.

In the above embodiment, the image sensor S1 is provided so as to image the field scene diagonally forward and downward. However, the image sensor S1 is configured to image the field scene vertically and not diagonally forward and downward. In this case, the plurality of long masks, for example, the pair masks M, as shown in FIG.
1 and M2 are arranged so as to be parallel to each other. The interval d between the pair masks M1 and M2 is set so as to correspond to the arrangement of the previously predicted rows of the seedlings Tn, T1 and T2 when the seedlings Tn, T1 and T2 are imaged in the proper setting state. , Each mask width Δd is set in each region T
The width is set to a width (for example, about 2 to 3 times the width of each area) that sufficiently covers the horizontal width of the screen of an, Ta1, and Ta2.

To explain the plurality of mask movement positions in this case, a mask center line G parallel to the longitudinal direction of the pair masks M1 and M2 is set at an intermediate position between the pair masks M1 and M2. Moving points moved by a set distance (XH / m) obtained by equally dividing the screen width XH by m from the screen center position (ε = 0) on the screen horizontal axis passing through the screen center, and the respective moving points At a predetermined angle (180 / n) obtained by equally dividing 180 degrees by n around the point where the mask center line G cuts the horizontal axis of the screen.
And the tilt angle γ when rotated and tilted. Therefore, the point at which the mask center line G crosses the horizontal axis of the screen corresponds to the specific point in this case. With respect to the angle, a direction along the screen vertical direction is used as a reference, and a counterclockwise direction from this direction is defined as a plus direction. FIGS. 10 and 12 show a flowchart of the control operation and the (ε, γ) table in this case. Basically, the ρ of FIG. 2 and FIG. Is changed to XH. In this case, the straight line L seen on the screen directly corresponds to the straight line L on the ground surface.
The conversion process from the imaging surface to the ground surface is unnecessary.

In the above embodiment, the seedlings Tn arranged in a row adjacent to the unplanted area N in the field determined based on the position information of the paired masks M1 and M2 at the mask movement position where the number of areas becomes the maximum. In order to obtain control information for steering control from the information of the area Tan (specifically, the area Tan located inside the left mask M1) corresponding to the left mask of the pair masks M1 and M2, Although the information of M1 itself was used, as another method, the line segment L connecting the obtained regions Tan is calculated as a straight line by, for example, a Hough transform process which is a straight line approximation means, or a method of least squares is used. And the like, information obtained by linear approximation or curve approximation can be obtained.

Here, a case where a straight line is approximated by the Hough transform will be described. As shown in FIG. 13, the x-axis passing through the center of the imaging field of view of the image sensor S1 is used as a reference line in the polar coordinate system, and the area Tan (actually, a plurality of pixels in the pixel density (32 × 32 pixels / 1 screen)) is used. ) Corresponding to the inclination θH set in advance in a plurality of steps in the range of 0 to 180 degrees with respect to the x axis based on the following equation (i) and the origin, that is, the center of the imaging visual field. And the distance ρH from the center of the screen. ρH = y · sinθH + x · cosθH (i)

Then, for one pixel, until the value of the inclination θH set in the plurality of steps reaches 180 degrees.
After repeating the process of adding the obtained two-dimensional histogram for counting the frequency of each straight line, the frequencies of a plurality of types of straight lines passing through each pixel are counted for every pixel. Then, when the counting of the frequency of the straight line for each pixel is completed, a combination of the gradient θH and the distance ρH, which is the maximum frequency, is obtained from the value added to the two-dimensional histogram, thereby obtaining one straight line having the maximum frequency. Lx is determined, and the straight line Lx is obtained as information obtained by linearly approximating a line segment L connecting the planted seedlings Tn adjacent to the unplanted area N on the imaging surface of the image sensor S1.

In the above embodiment, the seedlings Tn arranged in a row adjacent to the unplanted area N in the field are determined based on the position information of the paired masks M1 and M2 at the mask movement position where the maximum number of areas is reached. Line segment L corresponding to corresponding area Tan
Was obtained as the control information for the steering control. However, instead of obtaining the line segment, for example, the obtained area Tan
It is also possible to calculate the average position in the screen (four in FIG. 4) and obtain the control information from this position information.

In the above embodiment, a plurality of long masks, for example, pair masks M1 and M2 extending in a radial direction from the intersection VP on the front side in the traveling direction or pair masks M1 and M2 in a parallel state are connected to the intersection VP or the intersection mask VP, respectively. Although the point where the mask center line G cuts off the horizontal axis of the screen is set as the specific point and the mask is rotated around the specific point,
The specific point is not limited to the above. In the former case of the paired masks M1 and M2 extending radially, for example, the upper side or the lower side of the screen is rotated at a specific point, and in the case of the latter paired masks M1 and M2, for example, The specific point can be variously set such that a point at which the mask center line G cuts the upper side position or the lower side position of the screen is set as the specific point and rotated.

Further, in the above embodiment, a plurality of long masks are used as two pair masks M1 and M2 extending in a radial direction from the intersection point VP on the front side in the traveling direction or pair masks M1 and M2 in a parallel state. Although the mask is constituted by a mask, the number of masks constituting a plurality of long masks is not limited to two, and a seedling array imaged on the screen by the imaging direction of the imaging means S1 and the screen magnification and the like. 3 according to the number of
It is possible to increase and set more than the number. Also, the width of the mask can be set narrower or wider than that of the above embodiment. As exemplified above in various cases, the specific configuration of the calculating means 101 can be appropriately changed according to the situation of the device, the field, and the like.

In the above embodiment, a color image sensor S1 is used as an image pickup means for picking up an image of a field scene.
By subtracting the blue information B from the green information G and binarizing it based on the set threshold value, a region Tan, Ta1, Ta2... Corresponding to the seedlings Tn, T1, T2. However, for example, three primary color information R, G,
Using all of B, their ratio is the seedlings Tn, T1, T2 ...
The area within the set ratio range corresponding to the color
.. may be extracted as n, Ta1, Ta2,..., or the brightness signals of other locations such as seedling locations and mud surfaces may be set to appropriate threshold values by using a monochrome image sensor S1 as the imaging means. By binarizing based on
The area may be extracted by a simple device, and the specific configuration of the area extracting means 100 can be variously changed.

Further, in the above embodiment, the present invention is applied to an apparatus for automatically running a rice planting machine along a seedling row planted in a field, and an approximate line segment corresponding to a detected crop row is applied. Although the case where the information is configured to be used as control information for steering control has been described as an example, the present invention can be applied to an apparatus for detecting information on approximate line segments corresponding to various crop rows. The specific configuration of each unit, such as the type of crop and the configuration of the running system of the machine, and the use form of the information on the approximate line corresponding to the detected crop row can be variously changed.

Incidentally, reference numerals are written in the claims for convenience of comparison with the drawings, but the present invention is not limited to the configuration of the attached drawings by the entry.

[Brief description of the drawings]

FIG. 1 is a block diagram of a control configuration.

FIG. 2 is a flowchart of a control operation;

FIG. 3 is an explanatory diagram of image processing.

FIG. 4 is an explanatory diagram of image processing.

FIG. 5 is an explanatory diagram of image processing.

FIG. 6 is an explanatory diagram showing a relationship between a body traveling direction and an approximate straight line in an imaging visual field.

FIG. 7 is a schematic plan view of a rice transplanter.

FIG. 8 is a schematic side view of the same.

FIG. 9 is an explanatory diagram of image processing of another embodiment (1).

FIG. 10 is a flowchart of a control operation according to another embodiment (2).

FIG. 11 is an explanatory diagram of image processing of another embodiment (2).

FIG. 12 is an explanatory diagram of image processing of another embodiment (2).

FIG. 13 is an explanatory diagram of a Hough transform process according to another embodiment (3).

[Explanation of symbols]

Tn, T1, T2 crop S1 imaging means Tan, Ta1, Ta area Tan non-planting side area 100 area extracting means N non-planting area 101 arithmetic means M1, M2 long mask M3 crop presence detection mask VP intersection

Continuation of the front page (56) References JP-A-3-12713 (JP, A) JP-A-2-154606 (JP, A) JP-A-63-277908 (JP, A) JP-A-4-293403 (JP) , A) JP-A-4-325003 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) A01B 69/00 303 A01C 11/02 331 G05D 1/02 G06F 15/62 380

Claims (4)

(57) [Claims] Crops arranged in a plurality of rows (Tn, T1, T1)
Imaging means (S) for imaging a field scene in a predetermined range including 2)
1) and an area (Ta) corresponding to the crop (Tn, T1, T2) based on the imaging information of the imaging means (S1).
area extracting means (10, n, Ta1, Ta2)
0) and the non-planting side area (Tn) corresponding to the crop (Tn) adjacent to the unplanted area (N) in the field based on the area information extracted by the area extracting means (100). And a calculating means (101) for determining the crop (Tn, T1, T2) on an imaging screen of the imaging means (S1). ) Along the row direction and the crops (Tn, T1, T1).
2) has a predetermined width in a direction intersecting with the column direction, and the imaging means (S1) is in a proper setting state and the crop (Tn,
When the image of T1, T2) is taken, the crop (Tn, T1, T2) is taken.
T2) A plurality of long masks (M1, M2) arranged side by side so as to correspond to the predicted arrangement of the columns are set, and the plurality of long masks (M1, M2) are At each of a plurality of moving points separated by a set distance in a direction intersecting the row direction of the crop (Tn, T1, T2), a plurality of moving points specified as positions separated by a predetermined angle around the specific point in the row direction are set. The number of the regions (Tan, Ta1, Ta2) located inside each of the plurality of long masks (M1, M2) is counted at each of the plurality of mask movement positions. The area (Tan, Tan) among the plurality of mask movement positions.
The non-implantation side area (Tan) is obtained based on the position information of the plurality of long masks (M1, M2) at the mask movement position where the count value of Ta1, Ta2) becomes the maximum. Crop row detector. 2. The crop row detecting device according to claim 1, wherein the crop presence detection mask (M3) for detecting the presence or absence of the crop (Tn, T1, T2) is the plurality of long masks. At a position closer to the unplanted area (N) than (M1, M2), a column of the crop (Tn, T1, T2) predicted in advance for the plurality of long masks (M1, M2). The calculating means (101) is provided so as to have an arrangement relationship corresponding to the arrangement, and the region (Tan, Ta1, Ta2) is provided inside the crop presence / absence detection mask (M3).
Is present, the plurality of long masks (M
1, M2) The regions (Tan, T
a crop line detector configured to subtract a predetermined value from the count value for a1, Ta2). 3. The crop row detecting device according to claim 1, wherein the imaging means (S1) is provided so as to image the field scene diagonally forward and downward, and the plurality of elongate pieces. Mask (M1, M2) is used for the crop (Tn, T1, T1).
T2) A crop row detection device arranged so that a plurality of lines extending in the respective row directions forward in the traveling direction extend radially from an intersection (VP) where they intersect at one point. 4. The crop row detecting device according to claim 1, wherein the imaging means (S1) is provided so as to image the field scene vertically downward, and the plurality of elongated shapes are formed. A crop row detecting device in which the masks (M1, M2) are arranged in parallel with each other.