JPH0634229B2 - Imaging type boundary detection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of use] The present invention relates to an imaging means for imaging a boundary between an unprocessed work site and a processed work site on the front side in the vehicle body traveling direction in a two-dimensional direction, and an imaging device. Illuminating means for illuminating the imaging range of the imaging means,
Differentiating means for determining the differential value of each pixel based on the detected value of the brightness of each pixel arranged in the two-dimensional direction from the image pickup information of the image pickup means, in a state where the boundary is viewed obliquely from above, A binarization unit that compares the absolute value of the differential value of each pixel with a set threshold value to extract pixels at a location that is greater than or equal to the set threshold value, and pixel information extracted by the binarization unit corresponds to a boundary. And a line extraction processing means for performing line extraction.

[Conventional technology]

The above-mentioned imaging-type boundary detection device uses the unprocessed work site as the unprocessed work site in order to use the work vehicle as steering control information for automatically traveling along the boundary between the unprocessed work site and the processed work site. Since the brightness looks different from that of the processed work site, the fact that the brightness changes greatly at the boundary portion is used to obtain the binarized image information corresponding to the boundary.

Then, in order to obtain information on the boundary on the front side of traveling before the vehicle travels to that point, the boundary is imaged in a desired state from diagonally above.

By the way, in this type of imaging type boundary detection device,
Since the binarized image information corresponding to the boundary is obtained by utilizing the change in the brightness of the image information, if the imaging condition, that is, the brightness within the imaging field of view changes, even if the same portion is imaged, Since there is a risk that the change in brightness will be different and proper binarization may not be possible, an illumination unit such as a strobe light emitting device that illuminates the inside of the imaging field of view in synchronization with the imaging process is provided.

By the way, this type of imaging-type boundary detection device is desired to be unaffected by the weather because the work vehicle is generally used outdoors, so that it is constant regardless of the weather. It is desired to be able to image with the brightness of.

However, since the imaging field of view is a state in which the work site on the traveling front side is viewed from diagonally above, even if the image is illuminated with uniform intensity at the position where the illumination means is provided, on the ground, the farther the imaging distance is, The amount of illumination becomes lower than that at a closer location, and the obtained two-dimensional captured image becomes darker at a location at a longer capturing distance.

It is known that when the boundary between the uncut land and the already-cut land on the lawn is detected as the boundary, the already-cut land looks brighter than the uncut land.

Then, as shown in FIG. 2 (a), the uncut land (B) is located on the left side and the already cut land (C) is located on the right side of the screen, and the boundary
When (L) is imaged in the vertical direction, it is darker on the part where the imaging distance is farther on the screen, that is, the upper side is darker than the lower side, and the differential value is calculated for each pixel based on the brightness value. When it is found, the differential value becomes smaller toward the upper side, and if the entire screen is binarized with a constant threshold value, it is shown in Fig. 2 (b).
As shown in, there is a disadvantage that information corresponding to the pixel (indicated by a broken line in the drawing) corresponding to the boundary located at the upper side, that is, the far side is missing.

Therefore, in the past, for example, an illumination means with a large amount of light is provided,
It is considered that the illumination range is expanded beyond the imaging field of view so that the change in brightness within the imaging field of view becomes as small as possible.

[Problems to be solved by the invention]

However, in the configuration in which the illumination means with a large amount of light is provided and the illumination range is expanded beyond the imaging field of view, not only the illumination means becomes large, but also the device becomes expensive.

Also, regardless of the distance, so that the amount of illumination on the ground is constant,
It may be possible to change the illumination intensity of the illumination means in accordance with the distance of the imaging location, but this has the disadvantage of complicating the configuration of the illumination means.

The present invention has been made in view of the above circumstances, and an object thereof is to accurately obtain binarized image information corresponding to a boundary without using a special illumination means.

[Means for solving problems]

The characteristic configuration of the image-capturing boundary detection apparatus according to the present invention is such that the value of the threshold is smaller for each pixel so that a pixel corresponding to a portion with a longer imaging distance is smaller than a pixel corresponding to a closer portion. A threshold value setting means for setting a threshold value for binarization is provided, and its action and effect are as follows.

[Action]

Since the amount of illumination decreases as the distance to the pixel increases, and the brightness change decreases, the threshold value for binarization based on the magnitude of the differential value If the pixel at the farther place is set to be smaller than the pixel at the closer place, the accurate binarization can be performed regardless of the difference in brightness due to the perspective of the imaging distance.

〔The invention's effect〕

Since accurate binarization can be performed irrespective of the difference in brightness caused by the distance of the imaging distance, it is possible to use an illuminating unit that provides an appropriate amount of light at a location where the imaging distance is short. So,
The image information corresponding to the boundary can be accurately obtained within the entire imaging field of view without using a special illumination means.

〔Example〕

 Embodiments of the present invention will be described below with reference to the drawings.

As shown in FIG. 4, a pair of left and right front and rear wheels (1) and (2) are provided so that both steering can be operated, and a lawnmower (3) is suspended on the lower abdomen of the vehicle body. Therefore, a work vehicle (V) used for cutting grass and weeds is constructed.

The boundary (L) between the uncut land (B) as the unprocessed work site and the already cut land (C) as the processed work site on the front side in the vehicle traveling direction.
The image sensor (S ₁ ) as an image pickup means for picking up the image in the two-dimensional direction of It is installed.

Explaining the imaging field of view of the image sensor (S ₁ ), the worker (V) is in a state of being properly along the boundary (L), the boundary (L) is the lateral direction of the field of view. It is arranged so that it is located in the center.

In the figure, (13) is an illumination device such as a stroboscopic light emitting device that illuminates the imaging field of view with a set light amount in synchronization with the imaging operation of the image sensor (S ₁ ).

Explaining the structure of the work vehicle (V), as shown in FIG. 6, steering hydraulic cylinders (4) and (5) for individually operating the front and rear wheels (1) and (2) respectively. Further, a control valve (6), (7) for it is provided, and a hydraulic continuously variable transmission (8) that is switchable between forward and reverse and is capable of shifting forward and backward.
Is linked to the engine (E) and the variable speed motor
(9) is linked to the speed change arm (10) of the speed change device (8).

A rotation speed sensor for detecting the output rotation speed of the transmission (8)
(S ₂ ) is provided, and the work vehicle is based on the detection information.
It is configured so that the traveling distance of (V) can be detected.

Further, a steering position detecting potentiometer (R ₁ ) for detecting the steering position of each of the front and rear wheels (1), (2),
(R _2), and said transmission (8) potentiometer over the meter speed detector for detecting indirectly the vehicle speed by detecting the operation state of the (R ₃₎ is provided, their detection information, the image sensor (S ₁ ) information corresponding to the boundary (L) obtained based on the imaging information, and the rotation speed sensor
(S ₂ ) based on the detection information, the microcomputer controls the worker (V) to automatically run along the boundary (L) between the uncut land (B) and the already cut land (C) Utilization Control Equipment (11)
Is provided.

In the figure, (H) is a steering handle for boarding operation,
(R ₀ ) is a potentiometer for detecting the operation position, and (12) is a speed change pedal for boarding operation.

However, various means for image-processing the image pickup information of the image sensor (S ₁ ) and detecting information corresponding to the boundary (L) are configured by using the control device (11). become.

Explaining the automatic running of the work vehicle (V), as shown in FIG. 5, a square uncut land (B) surrounded by the already cut land (C).
The part from one side to the opposite side as one work process,
The image processing is performed so that the work vehicle (V) automatically travels along the boundary (L ₁ ) between the uncut land (B) and the already cut land (C) on the side along the length direction of the work stroke. Steering control will be performed based on the information detected at, with the end side of one work stroke (L ₂ ), that is, with reaching the opposite side of the uncut land (B),
By repeating the automatic turn toward the starting end of the next work stroke in the direction intersecting with the work stroke, it is possible to automatically perform the lawn mowing work within a predetermined range in the so-called turning style. .

Adding an explanation while describing the work of the control device (11),
With the start of traveling from the start end side of the work stroke, each time the vehicle travels a set distance based on the detection information of the rotation speed sensor (S ₂ ), the boundary detection processing described below is performed, and the detection information is Based on this, the steering control is performed so that the work vehicle (V) automatically travels along the boundary (L ₁ ).

Then, based on the detection information of the rotation speed sensor (S ₂ ),
Upon detecting that the distance within the set range has been reached for the distance corresponding to the length of one work stroke, the process of detecting the boundary (L ₂ ) on the end side of that work stroke is performed and detected. Based on the information, the end position relative to the current position of the work vehicle (V) is calculated, and as the vehicle travels to the calculated end position, based on the preset turn pattern, the next work stroke is calculated. Will turn towards.

However, it is determined whether or not the work is finished by determining whether or not the vehicle has traveled the preset number of strokes based on the number of turns and the like. If the work is not finished, the boundary detection process and the steering control based on the detection information are repeated.If the work is finished, the work vehicle (V) is stopped and the whole process is finished. become.

To explain the displacement of the vehicle body with respect to the boundary (L ₁ ) along the length direction of the work process, there is a displacement of the position toward the lateral direction of the vehicle body and a displacement of the inclination with respect to the direction of the boundary (L ₁ ). , They may be calculated as quantitative values,
Alternatively, only the deviation direction may be calculated.

If the steering control is further explained, the front and rear wheels (1),
In the parallel steering type in which (2) is turned in the same direction and the vehicle body is translated, the positional deviation in the vehicle width direction is corrected, and the front and rear wheels (1), (2) are turned in the opposite direction. By using the four-wheel steering system in which the vehicle turns sharply by using the four-wheel steering system, the misalignment is corrected. However, the steering amounts of the front and rear wheels (1) and (2) may be different so that the positional and orientational deviations can be corrected at the same time.

Next, each means for detecting the information corresponding to the boundary (L) will be described based on the flowchart shown in FIG.

That is, along with the start of the boundary detection process, the image sensor (S ₁ ) performs the image capturing process, and the captured image information is set to a preset pixel density (32 × 32 pixels). ) Is quantized according to. However, in synchronization with this imaging process, the illumination device (13) operates,
The image pickup field of the image sensor (S ₁ ) is illuminated with the set light amount.

Next, as shown in FIG. 3, the pixel (e) to be processed
The boundary (L ₁ ) is crossed based on the following formula (i) using a mask for 3 × 3 pixels covering 9 pixels including 8 pixels (a to d, f to i) adjacent to The process of calculating the differential value (SX) of the brightness change in the x-axis direction on the image that is the direction
This is performed for each pixel arranged in the two-dimensional direction.

SX (x, y) = (c + 2f + i)-(a + 2d + g) ... (i) However, in this differentiation, the boundary (L ₁ ) with respect to the work vehicle (V) is Based on which side it is located,
The sign of the differential value (SX) is discriminated, and only the value having one of the positive and negative signs is obtained.

That is, in the equation (i), since the brightness of the pixel located on the left side is subtracted from the brightness of the pixel located on the right side of the image, when the right side is brighter, the differential value (SX ) Is a positive value, and is negative when the left side is brighter.

The process of calculating the differential value (SX) of the brightness change in the x-axis direction corresponds to the differentiating means (100).

Then, by extracting only the pixels whose absolute value of the differential value (SX) is greater than or equal to a preset threshold value, the pixels at the location where the brightness change is large are extracted.

However, as described above, the amount of illumination by the illumination device (13) is
In the image, since the upper side corresponding to the farther distance in the image is smaller than the lower side corresponding to the closer distance, the brightness in the y-axis direction is reduced in accordance with the distance. (See Fig. 2 (a)).

Therefore, if binarization is performed with a constant threshold value, as shown in FIG. 2C, there is a possibility that the pixel (indicated by a broken line in the figure) corresponding to the boundary portion of the portion where the imaging distance is long may be missing. Therefore, the threshold value is changed according to the imaging distance, that is, the coordinate value in the y-axis direction.

In other words, when illuminated with a fixed amount of light, the amount of illumination is inversely proportional to the square of the distance. Therefore, based on the following formula (ii), the binary value is calculated according to the coordinate value (y) of each pixel in the y-axis direction. The threshold for conversion (Vth) is automatically changed.

However, in the above equation, Vref is the value of the appropriate threshold value for the pixel located at the lowest stage (y = 32) on the image at the shortest distance.

It should be noted that the reason why the coordinate value (y) in the y-axis direction is subtracted from 33 is that the coordinate value in the y-axis direction is set so as to increase downward with the uppermost stage being set to 1. The value is 1
In this case, this corresponds to the location with the longest imaging distance.

Therefore, in accordance with the coordinate value (y) in the y-axis direction, the threshold value (Vth) is automatically set to be smaller for pixels at a longer imaging distance, and each pixel in the vertical direction on the screen 2 is appropriately binarized according to the brightness of the image information, and as shown in FIG. 2B, the image information corresponding to the boundary portion of the portion having a long imaging distance is not lost. .

Therefore, this binarizing process corresponds to the binarizing means (101), and the process of changing the setting of the threshold value (Vth) based on the coordinate value (y) in the y-axis direction is It corresponds to the threshold value setting means (102) for setting the binarization threshold value for each pixel so that the pixel corresponding to the far imaging distance becomes smaller than the pixel corresponding to the closer distance. .

Then, as shown in FIG. 1 and FIG. 6, the straight line extraction processing means (103) uses, for example, Hough transform processing to determine the maximum frequency among straight lines passing through each binarized pixel. By extracting the one straight line, the information corresponding to the boundary (L ₁ ) is obtained.

However, when the boundary (L ₂ ) on the end side of the work process is detected, the change in brightness increases in the vertical direction on the screen, so in the process of calculating the differential value (SX), the y-axis The brightness value between pixels in the direction will be subtracted.

Then, from the position and the inclination of the obtained straight line, the work vehicle
The actual positional relationship between (V) and each boundary (L ₁ ) and (L ₂ ) is calculated, and steering control and turning position control are performed.

By the way, since the image sensor (S ₁ ) is attached in a state of imaging the work site on the front side of the vehicle body from obliquely above, each boundary (L ₁ ) obtained from the captured image information,
The position change of (L ₂ ) on the screen is inversely proportional to the distance to the actual positions of the boundaries (L ₁ ) and (L ₂ ) with respect to the work vehicle (V), and becomes smaller as the distance increases.

Therefore, in each of the calculation of the position of the boundary (L ₂ ) on the end side and the calculation of the deviation with respect to the boundary (L ₁ ) of the stroke side, the actual position for the work vehicle (V) is corresponding to the detected position on the screen. Will be converted to the position.

[Another embodiment]

In the above embodiment, the threshold setting means (102) sets the threshold for each pixel in the y-axis direction corresponding to the imaging distance, but the same threshold is used for every several pixels. May be simplified to Further, the result of calculating the threshold value corresponding to the coordinate value in the y-axis direction may be made into a table and stored, and the value may be read out corresponding to the coordinate value in the y-axis direction. Can be changed in various ways.

Further, in the above embodiment, a straight line corresponding to the boundary (L) is obtained from the obtained binary image information, and the traveling of the work vehicle (V) for lawn mowing is controlled based on the information of the straight line. Although the case has been described as an example, the specific configuration of each unit for utilizing the obtained binarized image information can be variously changed according to the configuration of the work vehicle to which the present invention is applied.

It should be noted that reference numerals are added to the claims for convenience of comparison with the drawings, but the present invention is not limited to the structures of the accompanying drawings by the entry.

[Brief description of drawings]

The drawings show an embodiment of an image pickup type boundary detection apparatus according to the present invention, and FIG. 1 is a flowchart of boundary detection processing, and FIG.
(A), (B), (C) are explanatory views of the image, FIG. 3 is an explanatory view of the differential processing, FIG. 4 is an overall side view of the work vehicle, FIG. 5 is an explanatory view of the work site, and FIG. The figure is a block diagram of the control configuration. (B) …… Unprocessed work site, (C) …… Processed work site, (L) ……
Boundary, (V) .... working vehicle, (S ₁₎ ... imaging unit, (SX) .... differential value, (13) .... illuminating means, (100) .... differentiation means, (101) ...
... binarization means, (102) ... threshold value setting means, (103) ... straight line extraction processing means.

Claims (1)

[Claims] 1. An image pickup means (S ₁ ) for picking up an image of a boundary (L) between an unprocessed work site (B) and a processed work site (C) on the front side in the vehicle body traveling direction in a two-dimensional direction, and Illuminating means (13) for illuminating the image pickup range of the image pickup means (S ₁ ) is provided in the work vehicle (V) in a state where the boundary (L) is viewed obliquely from above, and the image pickup means (S ₁ )
Differentiating means (10) for obtaining the differential value (SX) of each pixel based on the detected value of the brightness of each pixel arranged in the two-dimensional direction from the imaging information of
0) and the absolute value of the differential value (SX) in each pixel, and a set threshold value are compared with each other to extract a pixel at a location where the set threshold value is exceeded, and the binarization means (101). (101) is an image-capturing type boundary detection device comprising a straight line extraction processing means (103) for extracting a straight line corresponding to a boundary from the pixel information extracted in (101), where the threshold value is a portion where the imaging distance is long. An image-capturing boundary detection device provided with threshold setting means (102) for setting a binarization threshold for each pixel so that the pixel corresponding to .