JPS62122507A - Steering control apparatus of self-running working machine - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Industrial Field of Application] The present invention captures an image of the working ground condition in a predetermined range in front of the vehicle in the direction of movement of the vehicle body using an imaging means, and divides the captured image information into two images based on the average brightness difference. Boundary detection means detects the boundary between the untreated work area and the treated work area by converting it into a value, and the vehicle body automatically runs along the boundary based on the boundary detection result by the boundary detection means. . The present invention relates to a steering control device for an automatic traveling work vehicle, which includes a control means for steering the vehicle in this manner.

[Prior Art] In this type of autonomous work vehicle, the boundary detection means uses a photo-interrupter type or light reflection type sensitive sensor, a contact type sensor, etc. to detect the direction of travel. The present invention is configured to detect the boundary position along which the vehicle body should follow by sensing whether the state of the work area at a specific point, such as the previous or current position of the vehicle body, is an untreated work area or a treated work area. There was.

However, the above-mentioned conventional boundary detection means detects the deviation of one point of the work area from the boundary, and the detection result is immediately used as a control parameter for the steering, which causes the following disadvantages. was there.

In other words, when the working area is in poor condition and the boundaries are discontinuous, it is easy to lose sight of the boundaries to be followed, so processing such as averaging the boundary detection signals over a predetermined travel section is performed to prevent malfunctions. There is a need 1. As a result, there is an inconvenience that the control response becomes slow.

In addition, since it is possible to detect only the deviation from the boundary at the point where the vehicle is currently traveling or the point immediately before it, even if the detected boundary is discontinuous, the actual direction of the boundary is straight and the vehicle simply moves straight. Even in cases where this would be acceptable, unnecessary steering control is performed in an attempt to follow this discontinuous boundary, resulting in large meandering or control loss, resulting in poor straight-line driving performance. There was also.

Furthermore, there was also a structural drawback in that the state, such as the direction of the boundary, could not be detected until after the vehicle had actually traveled.

Therefore, the present applicant developed a boundary detecting means that can detect the shape of the boundary along which the vehicle body should travel without actually traveling along it.

We have previously proposed a means for detecting the boundary as a continuous line by binarizing image information obtained by capturing a predetermined range of a work area based on the average brightness difference (Special @ 1983-211637). issue).

[Problems to be Solved by the Invention] In the boundary detection means that detects boundaries by binarizing the image information, in order to perform co-value processing based on the brightness difference, it is necessary to If the difference in brightness from the area where the work has already been done is not clear, there will be a problem in that false detections will occur frequently. In order to substantially increase the brightness difference, for example, after processing is performed such as ignoring changes in brightness of the image information component corresponding to the direction of travel of the vehicle or averaging the images to remove fine brightness differences in advance. , the change in brightness between adjacent pixels of image information is differentiated, and the portion where the differential value is large is determined as a boundary.

Therefore, for example, if there is a part that looks dark in a treated work area, or a part that looks bright in an untreated work area,
The brightness change in this part becomes large, and information similar to the boundary is obtained. Especially when the same information as the boundary appears continuously in the same direction as the vehicle's traveling direction, it becomes inconvenient that it cannot be distinguished from the original boundary. .

By the way, since the automatic traveling work vehicle described above travels while performing work, it travels on untreated work ground adjacent to treated work land. In other words, there is a treated work area on either the left or right outer side of the vehicle body corresponding to the direction of movement, and there is an untreated work area on the front side with respect to the working device.

The present invention has been made in view of the above-mentioned circumstances, and its purpose is to effectively utilize the relationship between the moving direction of the vehicle body and the working area situation so that it is difficult to distinguish it from the boundary from the image information of the working area situation. The purpose of this method is to remove information in advance to detect accurate boundary plates.

Means for Solving the Problem] In order to achieve the above object, the steering control device for an automatic traveling work vehicle according to the present invention has a method for binarizing the imaging (surface image information) based on the average brightness difference. , means for calculating a differential value of a change in brightness of image information, means for classifying the sign of the differential value by 'l'lJ, means for detecting a vehicle body traveling direction, and vehicle body traveling direction detection by the traveling direction retrieval means. The present invention is characterized by comprising means for binarizing the image information based on the result, depending on whether the differential value change of only one of the codes determined by the differential value code discriminating means is equal to or greater than a predetermined value. Its actions and effects are as follows.

1 Effect] In other words, by limiting in advance the direction of change in brightness of the image information to be binarized based on the vehicle traveling direction and the sign (positive, negative) of the differential value of the change in brightness of the image information, unnecessary It removes information. In other words, from the relationship between the traveling direction of the continuous body and the working land situation, parts of the untreated working ground that are difficult to distinguish from the treated working land, and parts of the treated working ground that are difficult to distinguish from the untreated working land are removed in advance. This clarifies the original boundaries.

[Effects of the Invention] Due to the above characteristics, the following excellent effects have been achieved.

In other words, depending on the direction in which the vehicle is traveling, the image information is binarized only if the sign of the differential value of the brightness change is positive or negative, and the image information is brightened from the direction of the unprocessed area to the direction of the processed work area. Since the information in which the change in value increases is detected as a boundary, it is possible to remove noise components included in each of the unprocessed work area and the four processed work areas in advance. Therefore, it is difficult to eliminate boundaries and errors after binarization! Information that is easy to distinguish can be efficiently removed from the image information in advance, and as a result, accurate boundary information can be obtained.

[Example] Embodiments of the present invention will be described below based on the drawings.

As shown in Figures 5 and 6, @ theory (2) and (2)
A lawn mowing device (4) equipped with a disc-shaped cutting blade is suspended vertically in the middle of the vehicle body f+l, which is configured so that both rear wheels (3) and (3) can be steered. Unmown grassland showing the running course (B)
A sensitive sensor (A) is provided as a boundary detection means to detect the boundary (L) between the mowed field fc) and the mown field fc). 1 by controlling this seven? A lawn mowing vehicle is configured as an automatically traveling vehicle that can automatically travel along the boundary (L).

In order to configure the above-mentioned copying center height), the monitor camera (5) as an imaging means is used so that its i image ffl field is set at the target boundary (
L. ) in a predetermined range of grassland (Dl)
This monitor camera (5) is installed at the tip of the Senna instruction frame (6) extending downward and forward of the vehicle body (1).
), based on the average brightness difference between the two images, 21
By straightening, the deviation amount (X) and deviation angle (θ) of the detection boundary (L) with respect to the target boundary (Lo) K are determined, and B
The detection boundary (L) is located at the center of the field of view of the monitor camera (5) in the longitudinal direction with respect to the vehicle body (1) so that the ifi entry boundary (L) and the target boundary (Lo) coincide. In this way, the steering operation is controlled to correct the direction of travel.

Hereinafter, a block diagram of a means for detecting a boundary (L) by binarizing an image captured by the monitor camera (5) is shown in FIG. 1, an explanatory diagram of the image shown in FIG. 2, and Third
This will be explained based on the flowchart shown in the figure.

First, the image signals (So) captured by the monitor camera (5) are once recorded in the 7th record memo (January 7), and the control device (8) configured by a microcomputer is
The original image signal (So) stored in the frame memory (7) (shown in Fig. 2 (a)) is calculated by, for example, changing the brightness of the central pixel of the image into an area partitioned by a predetermined number of dots. By performing a process of directly replacing the average in i order, the averaged image is averaged to obtain a blurred averaged image (Sl) as shown in FIG. 2 (b).

By the way, as shown in FIGS. 5 and 6, the field of view (D) of the Itr monitor camera (5) is such that it looks down diagonally from above the vehicle body (1) to the work area below the front. fl), and the captured original image signal (So) has a pixel density of the portion far from the vehicle body (1) (upper O11, 1 in FIG. 2) of the vehicle body i1.
The area closer to 1 (lower side in Figure 2) is rougher. Therefore, if you uniformly average the image (QdZ), the upper part of the image (QdZ) will be averaged every week from the lower part, and as a result, the lower part will be noisy and the upper part will be unnecessarily blurred. Processing is performed so that the averaged density becomes coarser from the bottom to the top.

Next, by binarizing the portions of the averaged image (Sl) where the brightness changes are large, that is, the pixel portions where the brightness changes are large, as shown in FIG.
A binarized image (S2) is obtained in which only the boundary (L) between the area and the mown area (C) is bright and the pond area is dark. Regarding the relationship between the working area conditions, that is, in this example, since the vehicle is traveling on an untreated working area (B) with a treated working area (C) adjacent to the right side of the vehicle body (1), the averaged image ( In order to average SL), only the information for which a negative (-) differential value is obtained, where the sign of the differential value of the brightness change changes from dark to bright from the left side to the right side in Figure 2, is binarized. According to Figure 2 (B) middle (L
′) on the working area (B) where the differential value of the rough or sloughed area is positive (+), or on the treated working area (C). Weeds and small uncut areas are removed in advance as noise.

As shown in Fig. 4, the traveling direction of the car body f+1 is the same as that of the car body (1) and the treated work area ( Since the positional relationship with C) is reversed left and right every stroke, the sign of the differential value used to determine the change in brightness is switched between positive and negative at the end of the direction change, as shown in (a) in Figure 4. If the treated work area (C) is adjacent to the vehicle body + 1 + right side,
As shown in Fig. 2 (middle), only information on the differential value of negative (→) where the brightness changes from dark to bright is binarized in the horizontal direction, and as shown in Fig. 4 (b), the car body (1 ) On the left side is the treated work area (
C) is adjacent, the original image signal (
So′), positive (+) where the brightness change becomes light-dark.
This is to binarize only the differential value information in the horizontal direction to remove unnecessary noise components.

Then, after processing the boundary (L) detected by the co-valued image (S2) Vc to form a continuous line, for example, an approximate straight line, the vehicle body fl) follows it as shown in Fig. The deviation amount (X) and deviation angle (θ) of the detection boundary (L) from the power target boundary (Lo) are calculated.

Then, based on the amount of deviation (X) calculated by the calculation, the front and rear wheels (2) and (3) are steered in the same direction to move the vehicle body (11) along the parallel road tIJ without changing its direction. After that, based on the deviation angle (θ), the front and rear wheels +211 (31) are steered so that they are in relatively opposite directions to reach the target boundary (
L. ) so that the detection boundary (L) and the target boundary (Lo) coincide within the field of view of the moniker camera 5), that is, the heavy object (1) is aligned with the target boundary (Lo). It is controlled so that it follows.

Therefore, even if a discontinuous boundary, a locally curved boundary, a boundary that is unclear due to uneven grass, etc. is detected, it can be accurately corrected as a continuous boundary This makes it possible to control the steering without making unnecessary steering operations or causing the driving direction to deviate from the original boundaries.

In Fig. 1, (9)I(lO)r/'i is $6(
21ri (2+ and rear wheel +3) Hydraulic cylinder that actually steers I+31, (If), (+2
) are electromagnetic pulps that drive the hydraulic cylinders (9) and (10), respectively, and (1 (6) is a motor that operates the shift position of the continuously variable transmission (14).
2) is the front wheel (2) and rear wheel (3) + (
3) a potentiometer for detecting the steering amount and providing a feed puncture to the control device (8); (R3);
Similarly, is a potentiometer for detecting the shift position of the transmission device 114) and providing a 74-stroke to the control device (8).

By the way, the original image signal CS. ) is averaged by
Large changes in brightness due to the influence of reflected light from cut grass leaves, etc., and areas where the grass is small and bulges and uneven areas that are difficult to distinguish from mown areas (C) in uncut areas (B). This is to reduce the detection error of the boundary (L+) in advance.

The averaging process may be performed by using an optical filter for blurring instead of the arithmetic processing, or by simply using the camera (6) in a so-called out-of-focus state. Furthermore, for example, when using an optical filter, the focus filter is a so-called soft focus filter, so that the field of view gradually becomes soft focus from the top to the bottom. Changes in image density due to distortion of the field of view (D) of the ErJ camera (5) can be corrected at the same time.

Furthermore, in order to detect the traveling direction of the vehicle body (fl), it may be determined automatically by changing the direction, or it may be set manually.
By the way, the untreated work area (B
), or in a forward-reverse driving mode in which the vehicle repeatedly moves forward and backward by closing the width at each -stroke, the vehicle body (1) and the treated work area (C), i.e., the boundary (L).
Since the left-right positional relationship with the differential cage does not change, it is only necessary to set the sign of the differential cage once at the start of work.

In addition, in order to find the boundary (L), the two-direction image (S2
) can be processed using the least squares method or the like from each pixel (dot) that has been connected to the pixel (dot), but instead of binarizing, the differential value of the brightness change may be processed as a weighting coefficient.

Furthermore, in order to find the boundary (L) as a continuous line, it may be calculated using the coordinates of the image of the boundary line part by the least squares method or Hough conversion.

[Brief explanation of drawings]

The drawings show an embodiment of the steering control device for an automatic traveling work vehicle according to the present invention, FIG. 1 is a block diagram showing the configuration of the control system, and FIG. ) is an explanatory diagram of the image signal, Fig. 3 is a 70-chart showing the operation of the control device, Fig. 4 is an explanatory diagram of the positional relationship between the vehicle body and the boundary, and Fig. 5 is a plane showing the overall configuration of the lawn mowing vehicle. FIG. 6 is a side view thereof. (ri...vehicle body, (5)...imaging means, (A
)... Boundary detection means, (B)... Untreated work area,
(C) ... Treated work area, (L) ... Boundary. Agent Patent Attorney Osamu Kitamura Figure 1 Figure 4 wE5 Figure 6

Claims (1)

[Claims] An unprocessed working area (B) is obtained by capturing an image of the working area in a predetermined range in front of the vehicle body (1) using the imaging means (5) and binarizing the captured image information based on the average brightness difference. Boundary detection means (A) for detecting the boundary (L) between and the treated working area (C);
Based on the boundary (L) detection result by A), the vehicle body (1)
A steering control device for an automatic traveling work vehicle, comprising a control means for steering the vehicle so that the vehicle automatically travels along the boundary (L), wherein To convert the image information into a value, a means for calculating the differential value of the change in brightness of the image information, a means for determining the sign of the differential value, a means for determining the sign of the differential value, and a means for calculating the differential value of the change in brightness of the image information, a means for determining the sign of the differential value,
) Based on the means for detecting the traveling direction and the detection result of the vehicle traveling direction by the traveling direction detecting means, whether the differential value change of only one of the signs determined by the differential value sign determining means is greater than or equal to a predetermined value. Image information 2 depending on whether or not
A steering control device for an automatic driving work vehicle that is equipped with a means to convert it into a value.