JP2017099790A - Traveling object operation system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an operation system for a model vehicle, robot, etc. that enables an operator to perform operation with a feeling as if the operator rides on the vehicle or the like, and has a simple structure.SOLUTION: An operator performs operation using an operation device such as a handle, operation pedal, and the like; and an operation signal from the operation device turns to a radio signal to control a model vehicle or the like. The model vehicle or the like comprises "coordinate position and direction" detection means; on the basis of the coordinate position and direction orientation, a virtual camera position is determined. In addition to the model vehicle, each of many structures such as a road, building, and tree has prepared computer graphic data on its external view and has a detected position; for the external view, CG data is used together with position data. An imaged picture on CG external view data from a virtual camera is calculated and displayed as a picture at a driver seat.SELECTED DRAWING: Figure 1

Description

The present invention relates to a system for manipulating model cars, model robots, and the like. At the same time, the present invention relates to a system for virtually experiencing or training driving of a real car or a human-sitting robot.
Or, it relates to a game in which a model car or a model robot is controlled and played.
The person-seat robot is a robot that is a model robot that can be enlarged and operated by a person. In the present invention, the person-seat robot is present in a virtual space.
A generic term for model cars and model robots is used as a traveling object.

There is a radio control system as a system for manipulating model cars and model robots. When the control lever attached to the radio controller is moved, the operation is sent as a radio signal to the model car, etc., and the model car receives and decodes the signal and controls the steering motor and travel motor, etc., and the travel direction and speed Control etc.
The drawback of this radio control system is that the pilot cannot see the intuition for maneuvering because he / she steers the model car to be manipulated from a distance.
Also, since the maneuvering method itself is different from maneuvering a real car, even if you practice practicing a model car with a radio control, it will not be a practice to maneuver a real car, and you will enjoy the fun of maneuvering a real car I can't do that either.

The following attempt was made as an improvement plan. A video camera is installed in the driver's seat of the model car and the image is transmitted by radio waves. The operator feels as if he is riding a model car while watching the transmitted video on the monitor screen, and controls a maneuver similar to the real car and transmits the control signal to the model car.

However, when a model camera is loaded with a video camera as in this system, the model car grows and at the same time the driving course increases, and all model cars require broadband video communication. Many. I also want to see the side and behind the car, and I want to see 3D images, but it is difficult to increase the number of cameras.
Furthermore, in order to make the road image look authentic, it is necessary to create a real three-dimensional model such as a building or a landscape around the road, which further costs and takes time.

In addition to model cars, martial arts games and soccer games using radio controlled robots are also commonly used.
However, since the driver's point of view is an objective point of view, the maneuvering is difficult and the maneuvering cannot be realistic.

Japanese Patent Laid-Open No. 2001-187279 Japanese Patent Laid-Open No. 2004-298538 Japanese Patent Laid-Open No. 2014-187279

  An object of the present invention is to eliminate these drawbacks of the radio-controlled system, and it is possible to drive a model car as if driving a real car, or in combination with a model robot, as if the operator became a robot. The challenge is to make it fun to drive and provide it with simple equipment by using computer graphics technology.

In order to solve the problem, the radio control technology in the real space and the computer graphics technology in the virtual space are fused as follows.
Create a real field where the model car will run in real space. Then, a virtual field similar to the real field is created in the virtual space.
Next, when there is a model car on the real field, position detection means for at least two points of the model car are provided, and “representative position and orientation” are obtained from the obtained two positions, and “representative position and orientation” Based on the data, a virtual car is installed on the virtual field. A virtual camera is set at a position corresponding to the eyes of the driver of the virtual car.

Then, the image obtained from the virtual camera is exactly the scenery that should be seen from the driver's seat of the model car.
If there is a more suitable position than the driver's eyes, a virtual camera may be set there, and the scenery from there is displayed on the display.
The operator looks at the display and steers the model car as if he was in the model car's driver's seat or from a viewpoint that is convenient for driving.

When your car is viewed from the driver's seat, you can only see the hood, but if you change the viewpoint and set the virtual camera position behind your car, you will see an image that includes your entire car. Will be visible on the screen.

In order to change the travel route of the model car, various types of real obstacles are installed in the real field. At the same time, a virtual obstacle having the same contour line is placed on the virtual field.
Since the space between a certain obstacle and another obstacle becomes a route or road on which the model car travels, various travel routes are created.
If the model car goes out of the real field, the position detection function is lost, and in order to prevent this, it is generally necessary to enclose the real field with a field outer frame. This field frame is also a kind of real obstacle.
Since the relationship between the model car and the real obstacle is a phenomenon in the real space, either one or both will be pushed and moved according to the laws of physics if they collide with each other. Obviously, if you hit a stationary real obstacle, the model car will be bounced off.
Since the relationship between a virtual vehicle and a virtual obstacle is a phenomenon in a virtual space, it simply represents the change in position and orientation according to the physical law in the real space between the model vehicle and the real obstacle as an image faithfully. .

Fixed obstacles (buildings, vegetation, banks, fences, bulkheads, field outer frames, etc.) are installed as follows.
(Fixed type) Real obstacles are fixed and installed in the actual field with the “position and orientation” determined in advance. This doesn't move even if a model car hits it.
Then, the virtual obstacle is placed in the virtual field at the same “position and orientation”.

Mobile real obstacles (objects whose position and orientation change when they collide with other objects, balls used in games, etc.) are installed as follows.
A moving type real obstacle is provided with a detecting means of “position of at least two points” or “position and orientation”, or an obstacle of a rotating body is provided with a detecting means of only “one point position”. Installed anywhere.
The virtual obstacle is placed on the virtual field by obtaining the “position and orientation” from the detected “position of at least two points” or using the data of “position of one point” in the case of a rotating body.

Traveling objects (automobiles, robots, etc.) are installed as follows.
The model traveling object in the real space is driven by a radio control signal from a driver (person) or a computer and travels.
The model traveling object is provided with “at least two positions” detecting means. Find the “representative position and orientation” from the position data of the two points,
The virtual traveling object is set in the virtual space using the detected “representative position and orientation” data.

In general, in order to obtain the “position and orientation” of an object in real space, it is easy to provide a position detection function for two points on the object. This is because even if an obstacle has an arbitrary shape, if the positions of the two points are known, the position and orientation of the entire obstacle can be specified.
<Example> The position coordinates of the object are the position coordinates of the midpoint (representative point) of the two points. The direction of the object is the slope of the line connecting the two points.

Further, in the case of an obstacle such as a cylinder or a cone, the orientation data is not necessary, and the obstacle can be obtained by a position detection means only for the center point (one point).

As a method for detecting the position of a point, two (2) vertical magnetic field coils are installed on a model object. Below the actual field, a position detection plate with a large number of parallel lines in the vertical and horizontal directions is placed. There is a method in which the parallel lines are scanned vertically and horizontally, and the relationship between the electromotive force due to electromagnetic coupling with the coil installed on the model object and the scanning signal is measured and calculated. (Fig. 5)
As another method, a video camera is placed on the upper part of the real field, a model car or an obstacle is photographed, and the image is recognized and analyzed to detect “position and orientation”. (Fig. 16)
In some cases, there are methods of installing two point light sources on an object, detecting the position of the point light source from a video image, and detecting “position and orientation”.
There is also a method of drawing two small circles on an object and detecting the position of the circle.

Furthermore, considering the use of a robot to enjoy a game, it is extremely important to control the robot from the viewpoint of the robot.
For example, when shooting a ball in a soccer game, the robot, the ball, and the goal can be kicked so that the robot's feet, the ball, and the goal can be viewed from a straight line. Having a means for automatically determining the virtual camera position from the position leads to an interesting game.

By this invention, the radio controlled model car becomes a completely different product. First, the pilot's viewpoint was objectively viewed from a distance. However, according to the present invention, the pilot can be controlled from a subjective viewpoint, and the viewpoint can be switched while driving. An example of the perspective is
(1) Viewpoint of sitting in the driver's seat (2) Viewpoint of tracking his car from behind (3) Viewpoint of looking down his car from directly above (4) Viewpoint of looking behind the car (5) View from the distance Perspective to look around

In addition, various viewpoints are possible. If two virtual cameras are set for both eyes, stereoscopic viewing can be easily realized. In addition, using a head-mounted display with head orientation detection enables driving in a virtual world where all directions can be seen in stereoscopic view.
When viewed as a game, onlookers can see car races and soccer games of model cars in a real space without deception. (In video games, software can determine all phenomena, And I can't trust the results)
Or you can watch a video showing the entire car race in the virtual space on a large display. The pilot can then enjoy the thrill and pleasure of driving a real car while watching a 360 degree virtual space display.

  In addition, by bringing the driving performance of a model car closer to the real thing, it can be used for driving training. For beginners, it is difficult to understand the view from the driver's seat and the actual positional relationship of the car, but it can be understood by comparing the sight seen from the driver's seat with the display and the objective image of the model car that runs synchronously. Get early.

Note that the scale of the model car can be set arbitrarily by defining the height of the virtual camera as the height of the eyes of the driver. The scale of the actual field may be made according to the scale of the model car.
If a large model car is used, the actual field in the real space will increase proportionally, and if a small model car is used, the actual field will also decrease.
Then, since a human is maneuvering in the cockpit, the virtual camera is in the virtual driver's seat, and the virtual world of the video displayed there is recognized as a life-size world.

In addition to normal road driving, the model car may accidentally collide with the side walls of the road or collide with other model cars. Due to these phenomena, the model car changes its direction and speed. And this is taken over to the virtual space through the position detection of the model car, and appears to the operator as a display image.
Then, the pilot operates to avoid a collision in the virtual space reflected on the display. Even if a collision occurs due to a mistake in maneuvering, all the phenomena that should occur due to the collision occur in the real space governed by Newtonian mechanics, and the video of the virtual space resulting from this progresses without contradiction.

This is because the field in the real space and the field in the virtual space are related only at the position where the model car can contact or collide with an obstacle or another model car, and the same shape. It is required to be a thing.
This is shown by an actual example. In FIG. 19A, it is assumed that the bumper 120 of the model car 1 hits the partition wall 6 (point 125) beside the road and the traveling direction changes like the model car 1a with a dotted line in the real space, and the vehicle continues to run.
In other words, the shape design of model cars and obstacles in "real space" and the "model car and obstacle design" in "virtual space" are the parts that can come into contact with each other (such as bumpers and guardrails). Only the shapes need to be similar, and shapes such as the top of an obstacle that cannot be bumped do not have to be similar at all, and free shapes are allowed.
Since the tree 6 and the like in the virtual space shown in FIG. 19 (B) are unrelated to the collision of the automobile, they can be freely present in the virtual space even if they are not in the real world.

Also, regarding the surface color and surface drawing design that are not related to physical collision, it is not necessary to be similar in all parts.
Further, the shape in the virtual space can be drawn only by the outline of the shape, and the surface can be made transparent or translucent.
For example, if you make my robot transparent, draw only the outline, and place a virtual camera behind my robot, the position of the ball in front of my robot, my robot's legs, and enemy robot And the goal looks straight through my robot with a transparent outline, so you can shoot and pass accurately.

In the case of a car, if you make your car in the virtual space transparent, you can see the bumper and front wheels of your car from the driver's seat, and you can drive more accurately. In addition, driving becomes easier, and it can be made more fun as a game.

Since a distant view of a high part of the virtual space cannot be in contact with an automobile or the like, a distant view of any design can be used. You can enter anything from mountain views, buildings, and cloudy skies.
A distant view in the virtual space will be described with reference to FIG.
The distant view is not related to the position of the virtual camera, and the way to cut the distant view changes depending on the orientation of the virtual camera. Therefore, the distant view is a horizontally long panorama image (FIG. 20) rotated 360 degrees horizontally. It is a simple method to use an image 153 that is obtained by logically connecting the left and right edges of the image as an endless image, scrolling according to the direction of the virtual camera, and cropped at the camera angle of view as a distant view. . Further, as in the virtual space diagram of FIG. 22, the distant view 152 can be considered as a circular curtain with a large radius, and the virtual field 120 can be expressed as surrounding from a distance.
However, the figure actually seen by the operator is a video image displayed on the virtual camera 141 # 1 or 141 # 2 that is tracked after the virtual car 100 # 1 or 100 # 2.
Of course, as described above with reference to FIG. 9, the position of the virtual camera can be placed at the position of the driver's eyes or at various positions in addition to tracking the automobile.

In general, the real field in the real space is to make a thing as an actual work, so it is difficult to make a complicated thing.
FIG. 21 is an example of an automobile race, and FIG. 21A is a plan view and a front view of a real space. The four outer field walls 5 and several central partition walls 6 around (A) form a traveling course of an automobile, all of which are made of a plate material having a constant thickness.

Therefore, for the partition 6 in the real space, the outer wall 5 and simple obstacles, cut a flat plate with a thickness exceeding the height of the bumper 120 of the model car as shown in the plan view of the partition and the obstacle, and paste it on the actual field. It is made. (Reference drawing of real space: Fig. 21 (A) Real space plan view, front view)
On the other hand, partition walls and obstacles in the virtual field in the virtual space need to be beautifully and precisely designed. Since the finished product is simply CG data, no matter how complicated it is, if it is made in large quantities, the unit price per unit can be manufactured inexpensively. (Reference view of the virtual space: Fig. 22 (B) Plan view of the virtual space, front view Bank 121 50 122 60, Tree 123,)

An example in which a camera is used for position detection will be described.
16 is an example in which the camera 110 fixed at the upper center of the field is used for position detection. The model car 1 # 1 is provided with infrared light emitting elements 111 and 112 on the front and rear, and the camera 110 detects the light rays emitted from them, and detects the position and orientation of the model car.
Although it is possible to capture the shape of a model car with a camera and process this figure to calculate the position and orientation, it may take time.
At the initial time, when a plurality of automobiles are used, infrared light is emitted in accordance with an address command for automobile control, and the position of the light emitting element is determined by the camera.
Then, the infrared camera 110 detects the position of two points of the model car, and the CPU converts the position into the position and orientation, sets the virtual camera, performs image processing in the virtual space, and outputs the image to the display.

A front view of the car race (virtual space) in the computer is shown in FIG.
The difference between FIGS. 17A and 17B of the real space automobile race and FIG. 18 of the virtual space will be described.
Regarding the four field outer walls 5 surrounding the real space, a plate material of a certain thickness is used in the real space so that it can be easily created as described above. It shows to FIG. 17 (A) (B).
On the other hand, in the virtual space, as shown in FIG. 18, a large bank 121 having a completely different design is drawn instead of the partition wall 5 and a tree 123 is planted on the bank.

In the real space of FIG. 17B, the bumper 120 of the model automobile 1 is in contact with only the partition wall lateral surface 5a of the partition wall 5, and therefore only the portion having the same role needs to have the same shape.
This is the bank lateral surface 121a in FIG. 18 of the virtual space.
Since the partition horizontal surface 5a in FIG. 17B and the bank horizontal surface 121a in FIG. 18 are in a similar positional relationship, even if the other partition walls 5 and the bank 121 are different in shape, it is completely impossible to drive a car. There is no problem.

The same applies to the partition wall 5 in the central portion, and the partition side surface 5a in FIG. 17B and the middle bank side surface 122a in FIG. 18 are in similar positions. Does not happen.

As the driving itself is subject to the laws of physics, if the road is partially made of a slippery material, the tire will be slippery when it comes to that place, and it will require skills for maneuvering and will be interesting.

The scenery seen from the driver's seat is drawn with computer graphics, so the setting location, season, and weather can be changed for each lap, so various effects are possible.
Obstacles with a position detection function exist on the road, including in the CG image, and can be moved, so the orientation, position, and shape are changed by a motor to increase the difficulty and gameplay. Can also be used.

Since the model car itself is a physical phenomenon with no deception, it can be a global car race classified according to performance. In addition, the driver's seat can be designed like a real racing car, and if the scale ratio is large, a car race of about 200 km / h is possible. And since the way of maneuvering and scenery are close to real cars, it is possible to hold competitions where real car racers actually win. Furthermore, if game characteristics are incorporated, an extremely safe, thrilling and interesting device can be created.

Overall view of Example 1 with one model car The figure which looked at the real field of Example 1 from the side Plan view (A) and side view (B) of the virtual space of the first embodiment Overall view of Example 1 with two model cars Coil position detection principle diagram Illustration of converting 2-point position data into position and orientation data Internal block diagram of video computer 15 Illustration of virtual camera Virtual camera position description Stereoscopic illustration of virtual camera Example of smartphone or tablet terminal usage Rotating body type obstacle explanatory diagram Example of using a rotating body type obstacle Illustration of a rectangular parallelepiped obstacle Illustration of tunnel type obstacle Overall view of camera position detection Principle of position detection by camera Video of two-point position detection by camera Model car side wall collision diagram CG image of distant view Top view and front view of model car race The CG figure of FIG. Front view of foreground, middle, and distant view composite CG

FIG. 1 is an overall block diagram of an embodiment in which there is one model car and one driver.
FIG. 1 is roughly divided into two groups, a traveling unit 45 and a control unit 46.
The upper traveling unit 45 is a place where the model automobile 1 is driven. The actual field 12 is composed of the field outer wall 5 and the partition wall 6, and forms a course in which the model automobile 1 runs around as shown in the plan view of the figure.

The place where the model car 1 is traveling on the actual field 12 is depicted in the traveling part 45 of FIG. 1, and FIG. 2 is a side view of the AA ′ cross section viewed from the side.
A coil 3 and a coil 4 are installed vertically in the lower part of the model car 1 in FIG. 2, and an AC signal from the transmitter 33 is passed through the coil at appropriate time intervals to generate vertical magnetic fields M1 and M2. .
This magnetic field passes through the position detection plate 11 below the actual field 12 in FIG.
As shown in the upper part of FIG. 1, the position detection plate 11 is sized to cover the entire real field 12, and this principle structure can scan a number of orthogonal parallel conductors with switch circuits 31 and 32 as shown in FIG. It is configured as follows. (A real switch circuit is an electronic circuit.)
The position detection plate 11 and the coil position detector 13 are combined to operate so as to detect the position of each coil of the model automobile, and two coil position data are detected.
However, the coil position data is based on the detection of the planar position by the position detection plate 11. Considering the use of the position detection plate 11 of FIG. 5, the horizontal direction is the x axis and the vertical direction is the y axis. Also, since the height z-axis is not detected, the z-axis position is the height of the ground, and z = 0.
These two points of data enter the next position / orientation calculator 14 and are converted into position / orientation data.

Conversion from the position coordinates of the two points to the position and orientation data is performed as follows.
In FIG. 6A, it is assumed that the position of the front coil 3 is P (xp, yp, 0) and the position of the rear coil 4 is measured as Q (xq, yq, 0). . Next, in FIG. 6B, if the representative position of the automobile is the intermediate point between the coils 3 and 4, and the representative position coordinates of the automobile are S (xs, ys, 0),
(1) The representative point S (xs, ys, zs) is the midpoint of the position coordinates of the two coils.
xs = (xp + xq) / 2
ys = (yp + yq) / 2
zs = 0
It becomes.
(2) The horizontal posture angle θ is the direction of a line connecting the position of the coil 3 to the position of the coil 3.
Then, referring to FIG. 6B, θ = ATN ((yp−yq) / (xp−xq))
Since the actual field 12 is horizontal, the vertical attitude angle η is set to zero.
η = 0
In this way, the representative point (xs, ys, zs) and the posture angle (θ, η) are calculated.

The maneuvering unit 46 at the bottom of FIG. 1 is where the pilot 30 steers toward the control device 2 while viewing the display image 9.
The image of the display 9 in the control unit 46 is created as follows.
The video computer 15 uses computer graphics technology to create a virtual scene similar to the scene of the traveling unit 45 based on the object data stored in the memory of the CG data 14. This is shown in FIG. 3 (A) CG plan view and (B) CG front view.
Then, the position / orientation Pθk of the virtual camera that is assumed to be in the driver's seat of the car is calculated based on the position / orientation data Pθ1 of the model car 1, and an image that will appear in the virtual camera is drawn on the display 9. .
The contents of the CG data 14 in the memory are data necessary for the video computer 15 to replace the real world model object in the traveling unit 45 (upper part of FIG. 1) with a virtual object.

FIG. 3 shows a virtual scene model similar to the traveling unit 45 created by the video computer 15 using computer graphics technology.
The right example of the CG data 15 column in FIG. 1 is a detailed example of the data, and the video computer 15 uses this data to construct the virtual stereo model of FIGS. 3 (A) and 3 (B). Yes.
This is an example in which the design pattern as shown in FIG. 3 is pasted on the CG field outer wall 50 and the CG partition wall 60 of the CG data 15.
Based on the position / orientation data Pθ1 of the virtual model CG car # 1, the position / orientation data of the virtual camera is determined, and the image is calculated and created.

A real-world model object is usually a world where a real object in the real world is reduced to some scale. The virtual world created by the video computer 15 is created based on the scale size of the model. However, for the pilot, the size of the control device 2 and the size of the image on the display 9 are the basis of sensation to touch and see, so the image of the virtual world in the display is life-size real. It will be felt in the world of the size.

The operator 30 of the control unit 46 controls the control device 2 while looking at the display 9.
The control device 2 transmits this control content as a wireless control signal 41 toward the model vehicle 1.
In this way, the pilot 30 can steer the model car 1 as if he were driving a real car.

FIG. 4 shows an embodiment of a system similar to that of FIG. 1, except that there are two model cars and two pilots.
Note that this can be extended to a general system with three or more pilots.
FIG. 4 is composed of two groups, a traveling unit 45 and a control unit 46, as in the first embodiment of FIG.
However, unlike the first embodiment of FIG. 1, the traveling unit 45 has two model cars 1 # 1 and 1 # 2 running. There are two control units, the control unit 46 # 1 and the control unit 46 # 2, and the first operator 30 # 1 belongs to the control unit 46 # 1 and controls the model car 1 # 1.
The second operator 30 # 2 belongs to the control unit 46 # 2 and controls the model car 1 # 2.

Since the model car has two 1 # 1 and 1 # 2 cars running in the running unit 45, the model car has the function of detecting the position of the coil for two cars, and the coil position detector 13 and the position / attitude calculation unit 14 As a result, position and orientation data Pθ1 and Pθ2 for two units are output.
These data Pθ1 and Pθ2 are output to both of the two control units 46 # 1 and 46 # 2.

The first control unit 46 # 1 will be described.
CG data 15 # 1 includes (a) structural data and (b) design data required to draw two CG cars, [1] CG car 100 # 1 and [2] CG car 100 # 2. (c) Position and orientation data are installed.
Among them, the position / orientation data Pθ1 and Pθ2 are updated through the position / orientation calculator 14 after the position detection is continuously repeated in the traveling unit.

In CG data 15 # 1, [3] CG field outer wall 50, [4] CG partition wall 60, and [5] CG actual field 120 data are also installed as CG objects with fixed positions and orientations. ing.
The video computer 16 # 1 draws the entire scene based on the installed CG data 15 # 1.
Since the position and orientation of the virtual camera 110 used at that time are assumed to be installed in the driver's seat of the CG automobile 100 # 1, etc., the virtual camera 110 is virtually based on the position and orientation Pθ1 of the CG automobile 100 # 1. The camera position and orientation are calculated.
Since Pθ1 is the position and orientation of the model car 1 # 1 of the traveling unit 45, the image of the visual A # 1 that can be seen from the driver's seat of the model car is displayed on the display 9 # 1.
At this time, the position of the model car 1 # 2 running in front is Pθ2, which is detected, calculated and updated, and installed in the CG data 15 # 1, so that the video computer The image by 16 # 1 is reflected as 100 # 2 in the image of display 9 # 1.
While watching this video, the pilot 30 # 1 controls the control device 2 # 1, the radio control signal 41 # 1 is sent to the model car 1 # 1, and the model car 1 # 1 continues to drive. it can.

The same applies to the second control section 46 # 2.
Regarding the control unit 46 # 2, the virtual camera position and orientation is calculated based on Pθ2 instead of Pθ1, and the video by the virtual camera is displayed on the display 9 # 2. Maneuver device 2 # 2.

In this way, it can be seen that a powerful car chase can be enjoyed with the system of two cars and two pilots.
It can also be estimated that no matter how many car races there are cars, it can be dealt with by detecting the position of all model cars and expanding the system in the same way.

FIG. 8 shows an example in which the position and orientation Pθk of the virtual camera is obtained from the position and orientation Pθ1 of the CG car.
FIG. 8A is a diagram when the virtual camera is placed at the same plane position as the representative point S (xs, ys, 0) of the automobile as described above.
The height of the virtual camera is the height of the driver's eyes as shown in FIG. 8 (D). As zs = hc, the position of the virtual camera is S (xs, ys, zs).

If there is a driver's seat on the left side and the image is aligned with the driver's line of sight, it will be as shown in Fig. 8 (B). If the horizontal attitude angle of the car is θ, see Fig. 8 (C). The position of this virtual camera is D (xd, yd, zd).
However,
xd = xs + dd * sinθ
yd = ys + dd * cosθ
zd = hc
It is.
If there is a driver's seat on the right side, set dd <0.
In this way, by changing the camera position calculation formula, it can be used with any vehicle, including large trucks and buses.

Another example of the position of the virtual camera is shown in FIG.
105 in the figure is the normal position of the virtual camera (driver's view), and 101 in the figure is a virtual camera that reflects the back of the automobile, and plays the role of a so-called rearview mirror. This video can be seen if a dedicated display is placed behind the driver and the driver turns around.
Alternatively, it can be projected as a small rearview mirror image in the display that projects the front.

102 creates an image of tracking his car from the rear. This video makes it easy to understand the relationship between your car and the road.

Since 103 can be viewed from directly above the car, it is convenient for grasping the surrounding situation, such as in a garage. In this case, not a camera image but a plan view image without distortion can be cut and used only near the car.
These videos can be switched by the operator's operation.
In any case, the position and orientation of these virtual cameras are based on “the position and orientation Pθn of the CG automobile 100 # 1”, that is, based on “the detected position and orientation Pθn of the model automobile 1 # 1”. become.

FIG. 10A shows an example in which two virtual cameras 110 and 111 correspond to the left and right eyes to generate a stereoscopic image.
FIG. 10B shows an example of an apparatus for viewing a stereoscopic image. The display 9 switches between a left-eye image and a right-eye image every frame. The pilot is an example of viewing the display 9 using 3D glasses 120 whose left and right are alternately turned on and off in synchronization with the pilot.

FIG. 10C is an example of viewing another stereoscopic image, and uses a head mounted display 122 that allows the left and right eyes to directly view the left and right images.

Furthermore, if the head-mounted display 122 is provided with a head orientation sensor, the virtual camera orientation φ is changed according to the detection orientation φ as shown in FIG. It is also possible to look over all directions by changing the direction of.

An example in which the entire control unit 46 of FIG. 1 is configured by one composite electronic device such as a smartphone 29 is shown in FIG.
That is, an input signal from the wireless communication interface 22 is sent to the video computer 13 and the video created by the video computer 13 is displayed on the display 9 so that the handle, accelerator or brake displayed on the display 9 can be input with a finger. Thus, the controller 21 is configured, and an output signal from the controller 21 is communicated with the traveling unit 45 from the interface 22 by radio waves.
The control signal 26 controls the model car 1 through the transmitter 28.
The position data 23 from the coil position detector 13 of the traveling unit 45 passes through the interfaces 23 and 22 in reverse to become an output image by the video computer 15 and is displayed on the display 9 of the smartphone as shown in FIG.
The pilot 31 # 1 can control using only a small smartphone. In addition, the handle operation can be performed by tilting the smartphone left and right.

1 and 4 use a plurality of partition walls 6 attached to an actual field 12 as obstacles at a fixed position. An example of using a semi-fixed obstacle will be described with reference to FIG. .
In FIG. 12, a rotating body-shaped obstacle in which the cylinder 7 and the hemisphere 3 are combined is used. FIG. 12A is a plan view of the traveling field, and FIG. 12B is a side sectional view.
Since there is no need to detect the direction of the obstacle in the form of a rotating body, only one position detection coil 3 is mounted on the lower part on the center line.

And by changing the position where the obstacle 7 is placed, the travel path can be changed arbitrarily. Normally, the installation position of the obstacle 7 is determined before the game is started, so that the position of the obstacle 7 need only be detected once at the start of the game.
When the position of the obstacle 7 is detected, the video computer 15 stores the position, writes the obstacle CG image 70 in the image data and displays the perspective view on the display 9 assuming that there is an obstacle.
The figure further shows that two model cars 1 # 1, 1 # 2 travel between obstacles. On the display 9, images of model cars, obstacles, and surrounding partition walls reflected on virtual cameras placed based on the respective car positions are displayed.

For obstacles other than the rotating body, two (or more) position detection coils are used so that the posture (orientation) at the time of installation can be expressed. FIG. 14 shows an example of adaptation to a vertically long rectangular parallelepiped partition wall.
The model diagram is shown in FIG. As shown in the front sectional view, two position detection coils 3 and 4 are installed in the lower part of the front and rear of the model, and the antenna 21 and the coil drive circuit are connected to drive the position detection coil.
The intermediate point P between the two coil positions becomes the representative point of the obstacle, and the direction of the straight line drawn from the position detection coil 4 to 3 is the posture of the obstacle.
The appearance CG data is as shown in FIG. 14B, and the lengths Lk, Hk, Wk, etc. of the three sides of the rectangular parallelepiped are input as properties. In this example, the dimension of the model is the same as the dimension of the CG appearance, and the CG appearance is decorated like a brick structure only in the texture.
Similarly to the example of FIG. 12, the design of the upper part of the obstacle is arbitrary, and various decoration shapes are possible.

More complex structures can be defined as well. FIG. 15 shows an example of a tunnel. (A) is a front view of the model diagram, (B) is a side view of the model diagram, and (C) is a perspective view of the model diagram. The tunnel model is provided with two position detection coils 3 and 4 inside, and can detect a representative point and an attitude.
(D) is an outline CG diagram of the tunnel, and such a figure can be drawn on the display 9 based on the detected coil position data.
When the car enters the tunnel, the outside of the tunnel disappears from view and becomes an image of the inner wall, and the image of the car running inside the tunnel is shown on the display. When the model car enters the tunnel, it cannot be seen from the outside, but the position detection by the coil is continued, and the image running in the tunnel is normally created, so that the driver's driving is not hindered.

An example in which a video camera is used to detect the position of a model car or an obstacle is shown in FIG. The upper part is a perspective view of the traveling unit 45.
A bottom plate is affixed to the field frame 5 having a square frame shape as shown in the figure.
The obstacle 6 is created as follows.
Assuming that the thickness of the field outer frame 5 is h, obstructions 6A, 6B, 6C (arbitrary number) obtained by cutting out the same thickness h plate in the desired shape on the bottom plate as shown in the figure Paste in any position and orientation to make a real field 12.
The model car 1 is placed on the actual field 12 and is driven to run.
The model automobile 1 includes a position detecting circular figure or a light emitting element as shown in the figure before and after the vehicle body.
The height of the two circular figures or light-emitting elements is set so that the height from the ground is h as the height of the field outer frame 5 as shown in the figure.

FIG. 17A is a lateral cross-sectional view of the position detection camera 110 and the entire system.
As can be seen in this figure, the field outer frame 5 and the obstacles 6A and 6B are placed on the bottom plate.
6C and two position detection circles or light emitting elements 3 and 4 for the model car 1 are mounted, both of which have an equal height h from the bottom plate.
The position detection camera 110 is fixed with the lens vertically downward. The position detection camera 110 takes an image of the entire real field in view.
That is, if an ideal camera is used, the outer frame 5 and all the top surfaces of the three obstacles 6 and the two circular figures or the position detecting light emitting elements 3 and 4 for the model car 1 are Since they are all on the same plane, the image displayed on the light receiving plate of the camera placed in parallel with this should be similar and should be able to be photographed without distortion.
Actually, the image reflected on the light receiving film is distorted by the lens of the camera, but this distortion can be measured in advance to correct the image.

In addition, since the size of the outer frame 5 is determined to be a fixed size for each system, if the size of the camera image is changed and the size of the outer frame 5 is adjusted to the actual size, all the images are the actual size. Will fit.
Since both the outer frame 5 and the obstacles 6A, 6B, and 6C have the same shape on the upper and lower surfaces, the camera image corrected in the previous period can be considered as a figure based on the stomach. Further, since the two circular figures of the model automobile 1 or the positions of the position detecting light emitting elements 3 and 4 are also figures of the height h, it can be considered as a figure position with reference to the bottom plate.
That is, all the images of the camera are considered to be drawn with reference to the upper surface of the bottom plate.
For example, if only the upper surface of the obstacle model is colored in a special color, the camera image can be displayed only on the upper surface of the obstacle. Since all the upper surfaces have the same height h, they have a similar shape regardless of the distance from the center axis of the camera. Therefore, correct all dimensions correctly based on the rectangular image on the upper surface of the outer frame 5. Can do.

  If both the outer frame 5 and the obstacles 6A, 6B, 6C are determined at first, they are not moved. And since it can be measured over time, the computer scans along the shape, surveys the entire shape, measures the position and direction, puts it into computer graphics data, and creates CG data Insert into 15.

On the other hand, since the model car 1 moves around, it must be calculated every time the screen changes (usually 1/60 second).
To save computation time, detect and respond to two points. First, in the initial state when the power is turned on, the entire screen is searched to find two points of the car. Next, the position coordinates of the two points are calculated.
In the processing in the normal state, the position coordinates are not greatly different from the previous position, so the vicinity of the previous position is scanned to find two points. If found, the position coordinates are calculated.
The representative position and orientation are calculated from the position coordinates of the two points.
From the basic CG data of the car made in advance and the “representative position and posture” obtained this time, a CG 3D overall image is created. Further, the position and orientation of the virtual camera are obtained from the “representative position and orientation” of the car and the driver position data.
The image of the virtual camera is calculated from the virtual camera position and orientation and the CG stereoscopic image, and the image of the driver's seat is displayed on the display 9. The pilot then looks at it and steers it, and the radio steering signal moves the model car.
In this way, the steering system of FIG. 16 can operate.

In FIG. 16, obstacles 6A, 6B, and 6C create a part having an arbitrary external shape, and paste it on the bottom surface (actual field). In this example, the shape of each obstacle part is read from the image of the position detection camera 100 in the initial state, and the shape data is used as CG data.

FIG. 18 is an example in which an obstacle shape is read in advance as CG data by a computer.
The obstacle parts are marked with two or more marks as shown in FIG. 18 by circular marks for position reading or light emitting elements.
In some cases, the identification number code is also given by the color of the part or the barcode.
Then, reading by the computer is only reading of the position of two points and the part number, and only the representative position and orientation need be calculated, so that simplification and higher speed can be achieved.

  If the area of the actual field is large, divide it into several locations, install many position detection cameras so that they overlap a little at each location, perform calculation processing, and combine those information into one information Process.

Used in radio controlled model cars, robot model related manufacturing, sales, amusement parks, etc.
Can also be used in driving school.

1 1 # 1 1 # 2: Model car 2 2 # 1 2 # 2: Control device
3 4: Position detection coil 5: Field outer wall
50: CG field outer wall 6: Bulkhead
60: CG bulkhead
100 100 # 1 100 # 2: CG car 7: Rotating obstacle
70: CG rotating body type obstacle 8: Bulkhead type obstacle
80: CG bulkhead type obstacle 9 9 # 1 9 # 2: Display
10 10 # 1 10 # 2: Antenna
11: Position detection plate
12: Real field
120: Virtual field
13: Coil position detector
14: CG data
15 15 15 # 1 15 # 2: Video computer
16: Create virtual field
17: Memory for appearance CG data (appearance data for cars # 1 and # 2)
18: Recreate virtual cars # 1 and # 2 (change to new position and orientation)
19: Virtual camera position and orientation calculation means
20: Virtual camera image generator
21: Steering signal output
22, 23: Interface
25: Position data
26: Control signal
27: Transmitter
28: Handle
29: Pedal
30: Pilot
31: Y-axis scan switch
32: X-axis scan switch
33: X parallel line
34: Y parallel line
42: Timing signal
85: Tunnel-type obstacle
85CG: CG tunnel type obstacle
90: Coil drive circuit
100: Virtual camera (front driver eye position)
101: Virtual camera (rear driver's eye position)
102: Virtual camera (tracking position)
103: Virtual camera (rear tracking position)
104: Virtual camera (overhead position / posture position)
P (xp, yp): Coordinate position of coil 3
Q (xq, yq): Coordinate position of coil 4
SS (xs, ys): Intermediate point of P, Q or representative point θ: Object orientation δ: Virtual camera angle of view φ: Head mounted display orientation A A # 1 A # 2 Virtual camera angle of view

Claims (7)

In a control system in which a pilot maneuvering a traveling object traveling on a real field with a wireless control device,
On the “real field”, installing an actual obstacle surrounded by an outer wall with a height that the traveling object cannot get over;
On the "real field", the area where the actual obstacle is not installed is an area where the traveling object can travel,
The traveling object on the “real field” is provided with means for detecting “positional coordinates and orientation”;
Creating a virtual field of the virtual world corresponding to the real field in the computer;
A virtual obstacle having the same contour shape as the outer wall of the real obstacle is installed in a virtual world of the same “position coordinates and orientation” as the real obstacle,
Based on the “position coordinates and orientation” data of the detected traveling object, “position coordinates and orientation” for the virtual camera is calculated, and based on the “position coordinates and orientation” for the virtual camera, the computer Installing a virtual camera in the virtual world,
The virtual camera images the virtual world, and the captured image is an image effective for maneuvering the traveling object;
The video is shown on a display;
The traveling object steering system, wherein the pilot controls the wireless control device in a state where the display can be seen.   In Claim 1, if the traveling object corresponding to the virtual world of the traveling object is a virtual traveling object, the "position coordinates and orientation" is the same as the "position coordinates and orientation" of the detected traveling object. The virtual traveling object is installed in a traveling object maneuvering system.   In Claim 1, 2, the said real obstacle is provided with a means for detecting "position coordinates and orientation", and the detected object is a virtual obstacle as a corresponding object in the virtual world of the real obstacle. 3. The traveling object according to claim 1, wherein the virtual obstacle is installed on the virtual field with the same “position coordinates and orientation” as the “position coordinates and orientation” of the real obstacle. Piloting system.   2. The position coordinates and orientation detection means for a traveling object according to claim 1, wherein the "position coordinates and orientation" detection means is a "position of at least two points" detection means for the traveling object. 4. A traveling object steering system according to item 3.   Claims 1, 2, 3, and 4 are characterized in that the "position coordinates and orientation" detecting means for an actual obstacle is "at least two positions" detecting means for the actual obstacle. Item 5. A driving system for a traveling object according to any one of items 1, 2, 3, and 4. 6. The traveling object according to claim 1, wherein each of the virtual camera and the display is installed for each of the left eye and the right eye so that a stereoscopic image can be viewed. Steering system. 7. The display according to claim 6, wherein a head mounted display is used as the display, and means for detecting the orientation of the head of the operator is provided, and the orientation of the virtual camera is changed based on the orientation of the head. The traveling object steering system described.