JPH09168315A - Boundary detector, display device for running stage and running control device in working vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable the accurate judgment if the site is an untreated ground or a treated ground on an image screen, the accurate detection of the boundary between the untreated and treated grounds, the display of the running state in terms of the detected boundary on a screen and the automatic running of the working machine along the detected boundary. SOLUTION: A working vehicle V which runs along a boundary between an untreated ground and a treated ground is provided with an imaging means S1 for picking up an image around a boundary. Depending on the information of the image picked up by the imaging means S1, it is decided which side of the image on the screen in terms of the lateral direction of the machine body is an untreated ground or a treated ground as follows. For example, the image picked up is divided into two parts on the screen in terms of the lateral direction of the machine body, the ratio of the screen area corresponding to the treated site to the total screen area of each divided screen side is compared to each other, and the larger side is decided as the treated ground side. Based on the information regarding the decision and the information of the image, the line segment L corresponding to the boundary is determined by an image treating means. Besides the imaging means S1 and the image treating means, the machine body is provided with an image display means for displaying a running state and a running control means for controlling automatic running.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a work vehicle traveling along a boundary between an unprocessed work site and a processed work site, and an image pickup means for picking up an image of the boundary portion, and image pickup information of the image pickup means. Image processing for obtaining a line segment corresponding to the boundary based on work site identification information that identifies which side of the screen direction along the machine width direction of the captured image is the unprocessed work site or the processed work site side And a traveling state display device for displaying a traveling state with respect to the boundary detected by the boundary detection device, and traveling for automatically traveling along the detected boundary. Regarding the control device.

[0002]

2. Description of the Related Art In the boundary detection device for a work vehicle described above, for example, a planting work vehicle (rice planting) for planting seedlings in a stock unit as a crop adjacent to an already planted work site (corresponding to a treated work site) at set intervals. In order to obtain control information when driving a machine along the line between the planted seedling row, that is, the boundary between the unplanted work site (corresponding to the untreated work site) and the planted work site, Based on the information of multiple crop areas extracted for each seedling, the line segment corresponding to the boundary is obtained by linear approximation or curve approximation processing using, for example, Hough transform. The information (hereinafter, also referred to as work site identification information) that is necessary for the operation to identify which side of the current screen in the screen width direction along the machine width direction is the planted side or the unplanted side is, for example, the start of traveling. The operator manually sets the initial value at the time by looking at the field and then Turning at field end or the like with for reorientation by switching the initial value, had to be determined SundeUe side or not planted side (e.g., JP-A-6
No. 149362).

[0003]

However, in the above-mentioned prior art, it is troublesome for the operator to manually set the information of the planted side or the unplanted side at the start of traveling, and the angle at the time of turning after the start of traveling. Depending on the conditions, etc., there is a risk of making a mistake in switching the work site identification information, and as a result, the boundary between the planted work site and the unplanted work site may not be properly detected.

The present invention has been made in view of the above circumstances, and its purpose is, for example, a work in which an operator manually sets the work site specifying information at the start of traveling in order to solve the problems of the above-mentioned prior art. While not required, the information on the unprocessed work site or the processed work site side is accurately obtained even after the start of travel, and the boundary detection is properly performed, and the traveling state for the detected boundary is manually steered. This is to display on the screen so that the operator can judge, or to automatically drive the work vehicle along the detection boundary.

[0005]

According to the structure of claim 1, a picked-up image obtained by picking up an image of the boundary portion by the image pickup means on the side of the work vehicle traveling along the boundary between the unprocessed work site and the processed work site. Based on the information, it is specified which side of the screen direction along the lateral direction of the body of the captured image is the unprocessed work site or the processed work site side, and the work site identification information and the imaged image information of the imaging means. Then, the line segment corresponding to the above boundary is obtained.

Therefore, the information as to which side in the screen direction is the unprocessed work site or the processed work site side is determined from the captured image information itself, so that the worker can see the farm field at the start of traveling as in the conventional case. If there is no need to initialize the unprocessed work site or the processed work site side, and when switching is performed due to the change in direction accompanying turning, even after the start of travel, the angle conditions during turning, etc. Depending on the situation, there is no problem that accurate switching is not performed, and accurate information about the unprocessed work site or the processed work site side is obtained,
There is provided a work vehicle boundary detection device capable of appropriately performing boundary detection between both work sites.

According to the second aspect of the present invention, in the first aspect, the processed areas corresponding to the processed work sites are extracted based on the picked-up image information, and the picked-up screens are provided on both sides in the screen direction along the machine width direction. In each of the divided screens, the side where the area ratio of the processed area to the screen is large is determined to be the processed work site side, or the unprocessed area corresponding to the unprocessed work site is based on the captured image information. It is determined that the side having a large area ratio of the unprocessed area to the screen is the unprocessed work site side in each of the divided screens that are extracted and are divided into both sides in the screen direction along the machine lateral width direction.

Therefore, while extracting the processed area or the unprocessed work area in the screen, for example, by recognizing the shape or orientation of the processed area or the unprocessed area, the processed work area or the unprocessed work side is identified. In comparison, it takes a long time to recognize the pattern, and a relatively simple process of obtaining the area ratio of the processed area or the unprocessed area to the screen promptly and appropriately processes the processed work area or the unprocessed work area. The side can be discriminated, so that the preferable means of the constitution of claim 1 can be obtained.

According to the configuration of claim 3, claim 1 or 2
In the case where the boundary portion is not imaged in the imaged image at the proper image capturing position with respect to the airframe of the image capturing means, the image capturing direction of the image capturing means is the above until the boundary portion is imaged in the image captured by the image capturing means. The work site identification information is obtained based on a captured image in which the proper image capturing position is changed along the lateral direction of the machine body while including the proper image capturing position on the center side, and the boundary portion is imaged.

Therefore, since the orientation of the work vehicle is greatly inclined with respect to the direction of the boundary, the boundary between the unprocessed work site and the processed work site is imaged in the imaged image at the proper imaging position with respect to the machine body. Even if it does not exist, from the information that the imaging direction was changed until the boundary part was captured in the captured image,
It is possible to specify which of the screen direction along the machine width direction is the unprocessed work site or the processed work site side.
Alternatively, a suitable means having the configuration of 2 can be obtained.

According to the configuration of claim 4, claim 2 or 3
In, at each position in the screen direction along the machine longitudinal direction of the image pickup means, the processed region that is adjacent to the most unprocessed work site side in the screen direction along the machine lateral width direction is obtained, and in a row along the machine longitudinal direction. The line segment corresponding to the boundary is obtained by linear approximation so as to connect the processed regions that are lined up, or the most processed work in the screen direction along the machine width direction at each position in the screen direction along the machine longitudinal direction of the image pickup means An unprocessed area adjacent to the ground side is obtained, and a straight line approximation is performed so as to connect the unprocessed areas arranged in a line along the longitudinal direction of the airframe to obtain a line segment corresponding to the boundary.

Therefore, for example, if the information of all the processed or unprocessed areas in the image pickup screen is processed, it takes a long processing time, whereas the information of the processing object information is obtained by using only the information of the areas located near the boundary. While reducing the amount,
By linearly approximating a line segment corresponding to a boundary that is substantially linear in a relatively narrow range, boundary detection can be performed quickly and appropriately, and the preferred means of the configuration according to claim 2 or 3 is obtained. To be

According to a fifth aspect of the present invention, in the second, third or fourth aspect, the unplanted work site as the unprocessed work site is separated from the unplanted work site by a set planting interval in the machine body traveling direction and in the machine body lateral width direction. A plurality of crops are planted with the planting width separated, and a plurality of crop areas corresponding to each of the plurality of crops are extracted as processed areas. Based on this, the work site identification information is obtained, and the line segment corresponding to the boundary is obtained.

Therefore, the crop areas which are the processed areas are extracted as a plurality of areas which are separated surely at a set distance from each other, and the crop areas which are individually separated are connected to each other so that the line segment corresponding to the boundary is accurately formed. Since it can be obtained, for example, in the case of a deformation type region where the processed region is continuous, there is no problem that it is not easy to accurately obtain the line segment corresponding to the boundary based on the information of the processed region, 5. Particularly suitable for detecting boundaries of work vehicles for planting crops.
A suitable means of the construction of

According to the structure of claim 6, a picked-up image obtained by picking up a boundary portion on the same screen of the image display means on the side of the work vehicle and the unprocessed image obtained in claim 1, 2, 3, 4 or 5. A line segment corresponding to the boundary between the work site and the processed work site and an image showing the displacement of the line segment in the machine longitudinal direction are displayed.

Therefore, for example, as compared with the case where the operator judges the deviation state between the boundary and the front-back direction of the machine by looking at the actual work site, the operator determines the deviation between the boundary and the direction of the machine from the display image. A running state display device for a work vehicle can be obtained that can be easily and accurately determined and that can be manually steered so that the machine body travels in an appropriate state along the boundary.

According to the structure of claim 7, the boundary determined in claim 1, 2, 3, 4 or 5 is such that the work vehicle travels along the boundary between the unprocessed work site and the processed work site. The traveling device is automatically steered based on the displacement information in the longitudinal direction of the body corresponding to the line segment, that is, so as to cancel the displacement.

Therefore, for example, when the operator manually steers and the machine body is traveling along the boundary, the maneuvering operation is performed to perform other work (eg seedling replenishment in a planting work vehicle). Even if the hand is temporarily released from the steering wheel or the like, the machine body can be automatically traveled along the boundary, and thus a travel control device for a work vehicle having excellent workability can be obtained.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, a boundary detection device, a traveling state display device and a traveling control device for a work vehicle according to the present invention are run in a field while a rice planting machine (planting work vehicle) as a work vehicle is planting work. The case will be described with reference to the drawings.

As shown in FIGS. 1 and 2, a seedling planting device 2 for planting a seedling T as a crop in a field is movable up and down on the rear side of the body of a work vehicle V having front wheels 1F and rear wheels 1R. The work vehicle V is provided with a plurality of seedlings T at a set planting interval in the machine body advancing direction and a set planting width in the machine body lateral direction with respect to the unplanted work site as an unprocessed work site. Are planted to form an already planted work site as a treated work site.

A television camera S1 as an image pickup means for picking up an image of the boundary between the planted work site and the unplanted work site in which the plurality of seedlings T are planted in rows is provided on the front side of the work vehicle V.
This television camera S1 is attached to the tip of the support member 4 protruding toward the lateral outer side of the machine body so as to image the boundary portion adjacent to the lateral outer side of the machine body obliquely from above. The line segment L corresponding to the row direction, that is, the boundary of the seedlings Tn adjacent to the unplanted work site is the image pickup screen of the television camera S1 in the state where V is properly along the seedling row along the machine traveling direction, that is, the machine longitudinal direction. It is set so as to be in a state in which it coincides with a traveling reference line segment La that passes through the center of in the front-rear direction.

A plurality of work steps from one end side to the other end side of the field are set in a state of being arranged parallel to the machine lateral width direction. In each work step, the line segment L corresponding to the already planted seedling Tn is set. To match the travel reference line segment La on the screen,
The operator manually steers the work vehicle V while looking at the display screen of the TV monitor 13 described later, or the work vehicle V is automatically operated based on the deviation information of the line segment L from the traveling reference line segment La. Steering is controlled. Since the traveling direction of the work vehicle V with respect to the field is reversed every time the vehicle travels one stroke, and the position of the boundary with respect to the work vehicle V is horizontally reversed,
One television camera S1 is provided on each of the left and right sides of the work vehicle V, and the television cameras S1 on the side of use are switched and controlled as described below. In FIG. 1, the boundary position is the left side of the work vehicle V, and the image pickup information of the left TV camera S1 is used.

Explaining the structure of the work vehicle V,
As shown in FIG. 3, the output of the engine E is transmitted to each of the front wheels 1F and the rear wheels 1R via the transmission device 5, and the transmission state detecting potentiometer R3 for detecting the transmission operation state of the transmission device 5 and the potentiometer R3 are provided. An electric motor 6 for shifting the transmission 5 for shifting the transmission 5 and an accelerator pedal 17 for manual shifting are provided. In the figure, S2 is a distance sensor for detecting the traveling distance based on the output speed of the transmission 5.

Front wheel 1F as a traveling device which can be freely steered
Also, the rear wheels 1R are configured to be individually steered by hydraulic cylinders 7F and 7R, respectively, and steering angle detection potentiometers R1 and R2 for detecting steering angles of the front and rear wheels 1F and 1R, and a steering wheel for manual operation. 16, a steering wheel operation angle detection potentiometer R0 for detecting the operation angle of the steering wheel 16, and an electromagnetically operated control valve 8F for operating the hydraulic cylinders 7F, 7R so that the detected steering angle becomes a target steering angle. 8R and are provided. In addition, front and rear wheels 1F,
Parallel steering type that steers 1R in phase and at the same angle, four-wheel steering type that steers front and rear wheels 1F and 1R in opposite phase and at the same angle, and two-wheel steering type that steers only front wheel 1F. The three types of steering types can be selected and used by a changeover switch (not shown). However, when the vehicle automatically travels along each work stroke, the two-wheel steering type is selected.

Control device 12 utilizing microcomputer
The control device 12 is provided with the steering angle detecting potentiometers R1 and R2, the steering wheel operating angle detecting potentiometer R0, the shift state detecting potentiometer R3, the distance sensor S2, and the automatic traveling switch for starting the automatic traveling mode. Each signal from 15 is input, and the control device 12 outputs drive signals for the electric motor 6 for shifting and the control valves 8F and 8R. Then, in the automatic traveling mode in which the automatic traveling switch 15 is turned on, the control device 12 drives the electric motor 6 for shifting so that the transmission device 5 is in an operation state corresponding to a preset traveling speed, and In the manual traveling mode in which the control valves 8F and 8R are driven so that the front and rear wheels 1F and 1R have the set steering angle, and the automatic traveling switch 15 is turned off, the front and rear wheels 1F and 1R are operated by the steering wheel 16 The control valves 8F and 8R are driven so that the corners are formed. In the manual traveling mode, the operator operates the accelerator pedal 17 to shift gears.

As shown in FIG. 4, the reflected light from the seedling T or the like in the field scene is a polarizing filter for preventing the direct reflected light from the water surface or the like and the visible light for blocking the incident of the visible light. Optical filter 14 consisting of a cut filter
After being transmitted, it is incident on the television camera S1.
That is, from the TV camera S1, the seedling T in the field scene is displayed.
An infrared image of infrared light of the like is obtained as analog image information (see FIG. 9A). In addition, this image is
Since the television camera S1 images the front side of the field in a diagonally downward direction, the farther the image, the shorter the distance.

Next, the control configuration (G in FIGS. 3 and 4) for obtaining the line segment L corresponding to the boundary and the displacement of the airframe with respect to the line segment based on the image pickup information of the television camera S1 will be described. By processing the analog image signal from the TV camera S1, the seedling area T corresponding to the seedling T
The digital image signal in which the information of a is binarized (see FIG. 9).
(B), or a video signal processing unit 9 for outputting the perspective distortion removal processing for the image of (b)) and a pixel for which the digital image signal is preset. An image memory 11 that stores image data corresponding to the density (32 × 32 pixels / one screen), and the image memory 11
The seedling area Ta adjacent to the unplanted work site is discriminated from the seedling area Ta by processing the image data in the seedling area Ta.
An image processing unit that obtains the line segment L by connecting the lines (see FIG. 9D) and outputs control signals such as switching between the left and right television cameras S1 and an image capture signal to the video signal processing unit 9. And 10 are provided. And
The image processing unit 10 uses the line segment L with respect to the traveling reference line La.
Discriminating the displacement state and outputting the discrimination information to the control device 12. In the figure, 13 is a television monitor including a liquid crystal panel and the like for displaying a captured image of the television camera S1 and a processing result of the image processing unit 10.

From the above, by using the video signal processing section 9, a plurality of seedling areas Ta (corresponding to crop areas) corresponding to each of the plurality of seedlings T are generated based on the image information of the television camera S1. A crop area extracting unit 101 for extracting a processed area corresponding to a processed work site is configured, and a plurality of crop area extracting units 101 are extracted by using the image memory 11 and the image processing unit 10. Seedling area Ta
Corresponding to the work site specifying information and the boundary for specifying which side in the screen direction along the machine width direction of the image captured by the television camera S1 is the unplanted work site or the planted work site side based on the information The calculation means 102 for obtaining the line segment L to be processed is configured. That is, the line segment L is generated by the crop area extracting unit 101 and the calculating unit 102 based on the captured image information of the television camera S1 and the work site specifying information.
The image processing means GS is configured to calculate the work site identification information based on the imaged image information of the television camera S1.

Explaining the configuration for obtaining the work site specifying information, the calculating means 102 divides the image pickup screen of the television camera S1 into both sides in the screen direction along the machine width direction as shown in FIG. 9C. In each of the split screens GL and GR (divided into left and right in the figure), the side where the area ratio of the seedling region Ta to the screen is large (the left split screen G in the figure)
Since the area ratio in L is larger, the left side) is determined to be the planted work site side.

Then, as shown in FIG. 9D, the calculating means 102, at each position in the screen direction (vertical direction in the figure) along the machine longitudinal direction of the TV camera S1, the screen direction along the machine lateral direction. The seedling area Ta that is closest to the unplanted work site side is obtained, and the line segment L is linearly approximated by a Hough conversion process described below so as to connect the seedling areas Ta arranged in a row along the machine longitudinal direction. Ask.

Then, as shown in FIG. 9 (e), the image captured by the television camera S1, the obtained line segment L, and the line segment L in the longitudinal direction of the machine are taken by using the television monitor 13. Image display means for displaying the displacement, that is, the image showing the angle deviation θs with respect to the direction of the travel reference line La (the screen vertical direction) on the same screen is configured.
Therefore, the operator manually steers the work vehicle V by looking at the display screen of the television monitor 13. Further, using the control device 12, a control means 100 for steering control of the front and rear wheels 1F, 1R is configured based on the information on the displacement of the line segment L with respect to the machine longitudinal direction (the angular deviation θs). . That is, the steering control is performed so that the angle deviation θs becomes zero.

Next, the control operation by the controller 12 and the image processing means GS will be described with reference to the flow charts shown in FIGS. When the control starts,
In the main flow (FIG. 5), seedling row detection processing based on the information from the TV camera S1 is performed, then it is determined whether the vehicle is in the automatic traveling mode or the manual traveling mode, and the manual traveling mode performs the traveling state display processing. In the automatic traveling mode, the direction control process is performed based on the displacement (angle deviation θs) information.

In the seedling row detection process (FIG. 6), first, an image picked up from the television camera S1 is taken in every time the vehicle travels a set distance or every set time (FIG. 9 (A)). A filtering process for noise removal, a process of binarizing with a set threshold value to extract a crop region Ta corresponding to the seedling T, and a minute false crop region Ta ′ as shown in FIG.
To reduce the crop area / expansion processing, and further to perform the perspective correction processing to remove the crop area Ta shown in FIG.
Get the image. In the figure, eight crop areas T are arranged in three rows.
The state where a1-Ta8 is extracted is shown.

Next, the screen is divided into two parts on the left and right sides, and the area ratio of the seedling area Ta is calculated on each of the left and right screens. Specifically, FIG.
As shown in, the number of pixels (pixel value is “1”) corresponding to the seedling area Ta out of 32 vertical pixels × 16 horizontal pixels (512 pixels in total) is counted on each screen. Therefore, the area ratio is (the number of pixels of the seedling area Ta) / 512. Here, the sizes of the pixel numbers SL and SR themselves of the seedling area Ta are compared, and if the pixel number SL on the left screen is larger than the pixel number SR on the right screen, the left side of the screen is the planted work site. On the contrary, if the pixel number SR on the right screen is larger than the pixel number SL on the left screen, it is determined that the right side of the screen is the planted work site. Then, if the left side of the screen is a planted work site, the rightmost seedling area Ta in the lateral direction of the screen, that is, the pixel at the right end point of the planted work site is determined. The leftmost seedling area Ta in the direction, that is, the pixel at the left end point of the planted work site is determined.

Next, regarding the pixel discrimination of the above-mentioned end points, FIG.
The case of pixel detection of the right end point shown in 0 will be described according to the flow of FIG. 7. As shown in FIG. 10, there are eight seedling areas Ta1 to Ta8 in the screen, each of which consists of a pixel having a value of "1", and the pixel data other than this seedling area Ta is "0". . In the figure, with the upper left corner of the imaging screen as the origin (0, 0) of the coordinate axes, the x-axis along the right direction of the screen,
The y-axis is set along the downward direction of the screen. Here, since the unplanted work site is on the right side of the screen, that is, in the plus direction of the x-axis, as described above, on each of the 32 lines parallel to the x-axis with the coordinate value of y = 0 to 31, the left side of the screen is displayed. Pixel (x = 0)
If the pixel value (data) is sequentially determined in the right direction from and the pixel of the value “1” located on the rightmost side is detected, as shown in FIG. 11, the position term [y] of this pixel is the unplanted work site. It becomes the pixel position of the right end point of the planted work site adjacent to.

FIG. 8 shows the flow in the case of detecting the pixel at the left end point. In this case, the pixel on the right side of the screen in each of 32 lines parallel to the x axis with coordinate values of y = 0 to 31. When the pixel value (data) is sequentially determined from (x = 31) to the left and the pixel with the value “1” located on the leftmost side is detected, the position term [y] of this pixel is adjacent to the unplanted work site. It becomes the pixel position of the left end point of the planted work site.

Actually, in the above processing, the above x
Only in the determination of the pixel located on the most right or left side in each line parallel to the axis, the pixel is located on the planted side (eg, x = 11, y = 2 in FIG. 10) with respect to the end point adjacent to the unplanted work site side. In some cases, it is necessary to remove these pixels by using the difference in the distance from the left end of the screen, for example.

Next, the line segment L connecting the pixels at the end points of the planted worksite adjacent to the unplanted worksite is linearly approximated by the Hough transform process. First, as shown in FIG. 12, with the x-axis passing through the center of the image pickup screen of the television camera S1 as the reference line in the polar coordinate system, the crop region T in which the right end point has been detected is detected.
A plurality of straight lines passing through each pixel of a1, Ta2 and Ta3 are
Based on the following formula (i), it is obtained as a combination of the inclination θ set in a plurality of stages in the range of 0 to 180 degrees with respect to the x-axis and the distance ρ from the origin, that is, the center of the imaging screen.

[0039]

[Equation 1] ρ = y · sin θ + x · cos θ (i)

Then, for one pixel, after repeating the process of adding the two-dimensional histogram for counting the frequency of each obtained straight line until the value of the inclination θ set in the plurality of steps reaches 180 degrees, , The crop area Ta1,
The frequencies of a plurality of types of straight lines passing through each pixel of Ta2 and Ta3 are
When the counting of the frequency of the straight line for each pixel is completed for each pixel, the combination of the slope θ and the distance ρ that is the maximum frequency is obtained from the value added to the two-dimensional histogram to obtain the maximum frequency. It is assumed that one straight line Lx is determined, and that straight line Lx is a straight line approximation of the line segment L corresponding to the boundary on the screen. Finally, this straight line Lx
Is calculated with respect to the traveling reference line La that passes through the center of the screen in the front-rear direction and is determined as the angular deviation.

Therefore, in the direction control for automatically traveling the work vehicle V along the boundary (row of the planted seedlings T aligned in the machine traveling direction), the inclination angle θs of the straight line Lx with respect to the traveling reference line La is set to zero. The steering operation will be performed in a two-wheel steering manner so as to approach them. The TV camera S1
It is also configured to determine whether or not the end portion of the work process has been reached based on the imaging information of. Specifically, FIG.
As shown in FIG. 3, a plurality of seedling areas T lined up in the machine width direction
A line segment that connects each pixel of a in the lateral direction of the machine body is obtained by linear approximation (Hough conversion), and the distance P1, P2 from the screen upper end at each of the left and right end portions of the imaging screen of the linear approximation
Is obtained as position information corresponding to the distance to the end of the work process. When the average value of the distances P1 and P2 becomes equal to or larger than the set value, it is determined that the vehicle has traveled to the work stroke end portion. In the automatic traveling mode, when traveling to the end of the work process, the vehicle automatically stops after traveling the set distance.

[Other Embodiments] In the above embodiment, a television camera S1 is used as an image pickup means to obtain an imaged image by infrared rays in which visible light is cut, and the infrared image is binarized to correspond to a crop (seedling T). However, for example, a color type television camera that captures a color image based on the three primary color information R, G, and B is used, and the color difference between the crop and the background (mud) in the color image is extracted. The crop area Ta may be extracted by.

The image processing means GS obtains the work site specifying information for specifying the unprocessed work site side or the processed work site side. On the contrary to the above-mentioned embodiment, the image processing means GS performs the unprocessed operation based on the information of the image pickup device S1. An unprocessed area corresponding to the work site (unplanted work site),
That is, the area ratio of the unprocessed area to the screen is large in each divided screen obtained by extracting the area of the pixel (value 0) that is not the crop area Ta and dividing the imaging screen of the imaging unit S1 into both sides in the screen direction along the machine width direction. You may make it discriminate | determine that the side which is is an unprocessed work side. It should be noted that dividing the imaging screen into both sides in the screen direction along the lateral width direction of the machine body is not limited to being divided into two. Further, other than the configuration based on the area ratio as described above, the work site identification information may be obtained based on the captured image information.

Further, in order to obtain the line segment L by the image processing means GS, at each position in the screen direction along the longitudinal direction of the body of the image pickup means S1, contrary to the above embodiment, the screen along the lateral direction of the body is displayed. The line segment L may be obtained by linearly approximating the unprocessed regions adjacent to the most processed work site side in the direction and connecting the unprocessed regions arranged in a line along the machine longitudinal direction. .

Next, the orientation of the machine body of the work vehicle V is largely inclined with respect to the direction of the boundary, and the entire image pickup screen of the image pickup means S1 is an unprocessed work site (unplanted work site), or When the boundary is not imaged because it is all processed work (already planted work), the boundary cannot be detected by the image processing means GS. Therefore, means for solving such an image state is configured. That is, as shown in FIG. 14, the image pickup direction of the television camera S1 as the image pickup means is changed along the lateral direction of the machine body with the proper image pickup position (the position shown by the solid line in the figure) on the machine body being included in the center side. An electric motor 18 as an image capturing direction changing unit is provided, and the image processing unit GS captures the image of the TV camera S1 when the boundary portion is not captured in the image captured at the proper image capturing position of the TV camera S1. It is configured to operate the electric motor 18 so as to change the imaging direction from the proper imaging position along the lateral width direction of the machine until the boundary portion is imaged in the image. The fact that the boundary portion is not captured in the captured image means that the information amount of the seedling area Ta in the screen is less than the lower limit set amount (all unplanted work sites).
Or conversely, it is judged whether it is larger than the upper limit set amount (all are in the state of planted work sites).

To be more specific, the television camera S1 is rotated about the longitudinal axis of the television camera S1 which is orthogonal to the image pickup direction (obliquely forward and downward) and is inclined forward in the plane along the longitudinal direction of the machine body. Then, as shown in FIG. 15, after the image G1 at the proper setting position with respect to the machine body, the image is changed to the left side first, and when the boundary portion is not imaged on the left side image G2, the image is changed to the opposite right side. Then, the image G3 on the right side is picked up (note that these three images G1, G2, G3 are
The rotation amount is set so as to be continuous in the lateral direction). Here, in the case of FIG. 15A, the image G1 at the proper setting position and the image G2 at the left side position are all planted work sites, but the boundary is imaged at the image G3 at the right side position. In the case of FIG. 15B, the image G1 at the proper setting position is displayed.
Are all unplanted work sites, but the boundary is captured in the image G2 at the left position, and the image G3 at the right position is not captured. Then, from the information of rotating the imaging direction from the setting proper position to the left or right until the boundary is imaged,
The position of the boundary with respect to the airframe is determined. Further, as for the image in which the above-mentioned boundary is captured, the above-mentioned (Fig. 9 (c) (d))
As shown in (2), the screen is divided into left and right, and a process for determining which of the left and right is the planted work site is performed.

In the manual traveling mode, since the boundary is not imaged on the screen of the television camera S1,
For example, in the case of the state of Fig. 15 (a), "the boundary is on the right side" is displayed, such as information on whether the boundary is on the left or right side of the screen, and steering information for traveling the aircraft along the boundary. .
Turn the steering wheel to the right. The instruction such as “” is displayed on the television monitor 13. Further, in the automatic traveling mode, the front wheels 1F are automatically steered so that the orientation of the machine body is changed to the side where the boundary is located.

In the above embodiment, the work vehicle is a planting work vehicle, and a plurality of crops T are planted on the unplanted work site as the unprocessed work site, and the planted work site as the treated work site. However, in addition to this, a harvesting machine for cutting crops, grass, etc. grown on an uncut work site as an unprocessed work site to form an already-cut work site as a treated work site. It may be a reaping work vehicle such as a lawn mower or a cultivating work vehicle that cultivates a pre-cultivation work site as an untreated work site to form a cultivated work site as a treated work site.

In the case of a mowing work vehicle, the untreated area corresponds to the area of crops and grass, and the treated area corresponds to the area where no crop or grass exists, contrary to the planting work vehicle. The extraction process is performed. In the case of a tillage vehicle, the untreated area corresponds to the uncultivated working area, the treated area corresponds to the cultivated working area, and both areas are uncultivated working areas. Since the light reflectance in the area is higher than that in the cultivated work area, the extraction processing is performed separately.

In the above embodiment, the image processing means GS uses the Hough transform for linearly approximating the line segment L. However, other than the Hough transform, for example, the least square method or the like is used. Good. The line segment L may be obtained by curve approximation instead of linear approximation.

In the above embodiment, the displacement .theta.s which is displayed on the image display means 13 as the information for the manual steering or is used as the steering control information for the automatic traveling is the line corresponding to the boundary in the longitudinal direction of the machine body. It was calculated as the angle deviation of the minute L,
The positional deviation in the lateral direction of the machine may be obtained together with the angular deviation.

The traveling device may be, for example, a crawler traveling device other than the wheel type traveling device as in the above embodiment.

In the claims, reference numerals are provided 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 schematic plan view of a planting work vehicle.

FIG. 2 is a schematic side view of the same.

FIG. 3 is a block diagram of a control configuration.

FIG. 4 is a block diagram of a control configuration.

FIG. 5 is a flowchart of a control operation.

FIG. 6 is a flowchart of a control operation.

FIG. 7 is a flowchart of a control operation.

FIG. 8 is a flowchart of a control operation.

FIG. 9 is an explanatory diagram of image processing.

FIG. 10 is an explanatory diagram of image processing.

FIG. 11 is an explanatory diagram of image processing.

FIG. 12 is an explanatory diagram of Hough transform.

FIG. 13 is an explanatory diagram of end detection at the end of the work process.

FIG. 14 is a schematic plan view showing a changed state of the imaging direction in another embodiment.

FIG. 15 is an explanatory diagram of image processing according to another embodiment.

[Explanation of symbols]

 S1 imaging means L line segment GS image processing means 18 imaging direction changing means T crop Ta crop area 101 crop area extracting means 102 computing means 13 image display means 1F, 1R traveling device 100 control means

Claims (7)

[Claims] 1. An image pickup means (S1) for picking up an image of the boundary portion on a work vehicle traveling along a boundary between an unprocessed work site and a processed work site, and imaged image information of the image pickup means (S1). A line segment (L) corresponding to the boundary is determined based on work site identification information that identifies which side of the screen image in the screen width direction of the captured image is the unprocessed work site or the processed work site side. A boundary detection device for a work vehicle provided with a desired image processing means (GS), wherein the image processing means (GS) determines the work site identification information based on the imaged image information of the imaging means (S1). A work vehicle boundary detection device configured to determine. 2. The image processing means (GS) extracts a processed area corresponding to a processed work area or an unprocessed area corresponding to an unprocessed work area based on the information of the image pickup means (S1), In each divided screen obtained by dividing the image pickup screen of the image pickup means (S1) on both sides in the screen width direction along the machine body width direction, the side where the area ratio of the processed region or the unprocessed region to the screen is large is the processed work site or the unprocessed work site. The work vehicle boundary detection device according to claim 1, wherein the work vehicle boundary detection device is configured to perform a determination process of determining that the processing work site is located and obtain the work site identification information. 3. An image pickup direction changing unit (18) is provided for changing the image pickup direction of the image pickup unit (S1) along the lateral direction of the machine body with the proper image pickup position with respect to the machine body being included in the center side. The means (GS) captures the boundary location in the captured image of the image capturing means (S1) when the boundary location is not captured in the captured image at the proper capturing position of the image capturing means (S1). Until the image pickup means (S1)
3. The work vehicle boundary detection apparatus according to claim 1, wherein the image pickup direction changing means (18) is configured to operate so as to change the image pickup direction from the appropriate image pickup position along the lateral width direction of the machine body. 4. The image processing means (GS) is adjacent to the most unprocessed work site side in the screen direction along the machine width direction, at each position in the screen direction along the machine longitudinal direction of the image pickup means (S1). The processed area or the unprocessed area adjacent to the most processed work site side is obtained, and the processed area or unprocessed area arranged in a line along the longitudinal direction of the machine is linearly approximated to connect the line segment (L ) The work vehicle boundary detection device according to claim 2 or 3, which is configured to obtain 5. The plurality of work vehicles are arranged in such a manner that a set planting interval is set in a machine advancing direction and a set planting width is set in a machine lateral direction with respect to an unplanted work site as an unprocessed work site. A planting work vehicle for planting a crop (T) to form an already-planted work site as a processed work site, wherein the image processing means (GS) is based on the imaged image information of the imaging means (S1). A crop area extracting unit (101) for extracting a plurality of crop areas (Ta) corresponding to a plurality of crops (T) as processed areas, and a plurality of crop areas extracted by the crop area extracting unit (101). Boundary of the work vehicle according to claim 2, 3 or 4, further comprising: computing means (102) for obtaining the work site specifying information and the line segment (L) based on the crop area (Ta) information. Detection device. 6. A picked-up image of said image pickup means (S1) picking up an image of said boundary portion, and claim 1, 2, 3, 4 or 5.
The line segment (L) obtained by the boundary detection device for the work vehicle described
And a traveling state display device for a work vehicle, which is provided with image display means (13) for displaying on the same screen an image showing the displacement of the line segment (L) with respect to the machine longitudinal direction. 7. A traveling device (1F, 1R) which can be steerably operated
And the traveling device (1F, 1R) based on the displacement information of the line segment (L) in the machine longitudinal direction obtained by the boundary detection device for a work vehicle according to claim 1, 2, 3, 4 or 5. A traveling control device for a work vehicle, which is provided with a control means (100) for steering control.