JPH031217A - Stereographic image processor - Google Patents

Abstract

PURPOSE:To perform a stereographic image processing according to the intention of an operator by controlling both cursor display positions for left and right eyes by reacting according to first input information to specify the position of a drawing in an in-plane direction and second input information to specify the position in a depth direction simultaneously. CONSTITUTION:Figure (A) shows states where a liquid crystal shutter 5a at a left eye side is set at a transmitting state and a liquid crystal shutter 5b at a right eye side is set at a shielding state, and furthermore, since a point P is displayed on a minitor plane 2a and a cursor for left eye at a perspective point L by the visual point of the left eye, an observer 10 observes the cursor at the point L only with the left eye. Meanwhile, figure (B) shows that the cursor at a point R is observed with only the right eye adversely in figure (A). Then, the cursors at the points L and R are displayed on a monitor 2 alternately with a time division system as images for left and right eyes, and the switching of perspection and shielding of liquid crystal glasses 5 is performed interlocking with the above display. In such a way, the observer 10 can observe a stereographic designation point P as a stereographic image with the cursors at the points L and R.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a three-dimensional image processing device, and more particularly, to a three-dimensional coordinate input device using a two-lens three-dimensional image system.

[Conventional technology] In recent years, image technology has developed at a remarkable rate, and in particular, three-dimensional image processing technology has come to be used in a wide variety of fields as the computing power of large computers to small personal computers has improved. . There is a demand for a stereoscopic image input device that is easy to use, and a two-dimensional image input device as shown in FIG. 13 has been put into practical use as a conventional coordinate input device. As shown in the block diagram of FIG. 14, the configuration of this device includes a controller 61, a CP061a that controls all elements, a ROM section 61b that stores a control program for this device, and a color display unit 62 that displays on a monitor 62. A color palette storage unit 61c that provides instructions for color palettes, a character storage unit O1d for character and line shape information to be similarly displayed on the monitor 62, and a VRAM that stores monitor display data.
81e, a keyboard switch 61f that gives control instructions to the CPU 61a, and an SS which is a video synchronization signal generation circuit for clock and horizontal and vertical synchronization.
The counter 61h controls the monitor display based on the output of the G61g6 and the SSG61g. On the other hand, a light pen 63 for specifying the coordinates of a designated point to the controller 61 is comprised of a photodetector 63a that detects the rask light from the monitor 62 and a designation switch 63b that is operated at the designated point. To explain the operation of this device, as mentioned above, this device is for two-dimensional use, and the point indicated with the light pen 63 is displayed as a cursor depending on the color and line selected with the keyboard switch 61f, and , locus data of line segments connecting the input points are stored in the VRAM 61e and displayed on the monitor at the same time.

Furthermore, as an input device for three-dimensional coordinates for a three-dimensional image, the device disclosed in Japanese Patent Publication No. 7514/1980 displays a left-eye image and a right-eye image of a three-dimensional image taken from different angles, In the means of performing stereoscopic vision,
This is a three-dimensional image observation device characterized in that a pair of left and right markers specifying the same three-dimensional point are configured to move in conjunction with each other on the image, and also to move and hang independently.

[Problems to be Solved by the Invention] However, in order to input stereoscopic image coordinates using the two-dimensional coordinate input device shown in FIG. It is not practical as it requires input from other sources, and the images are flat and cannot capture a three-dimensional effect. In addition, special public service 1977-
According to the method disclosed in Publication No. 7514, when specifying the coordinates of the three-dimensional image, if the depth is not changed, it is sufficient to keep the relative positions of the markers on the left and right images constant and move them on the left and right images in a linked manner. , it is possible to easily specify the coordinates, but when changing the depth, the input mode is changed by operating a switch when inputting, and the relative position of the marks on the left and right images indicating the coordinates of the 3D images on the left and right images is changed. However, specifying three-dimensional coordinates at will is difficult to operate and requires considerable skill.

In order to solve the above-mentioned problem, the present invention facilitates the operator's intention by inputting the center point and depth of a transparent point or a projection point on the display screen with a simple operation regarding the input of the coordinates of a stereoscopic image. An object of the present invention is to provide a three-dimensional image processing device that is capable of performing three-dimensional image processing in accordance with the following.

[Means and effects for solving the problem] In the stereoscopic image processing device of the present invention, both the left-eye cursor display position and the right-eye cursor display position in the two-lens stereoscopic image processing are in the in-plane direction of the screen. It is characterized in that it can be controlled in response to the first input information specifying the position and the second input information specifying the depth direction position of the same screen simultaneously.

[Example] First, before explaining the details of the example of the present invention, the concept of the present invention will be explained using FIGS. 1 to 4.

FIG. 1 shows an example of a stereoscopic image processing device using the two-lens stereoscopic image observation method of the present invention, and this device includes a controller 1,
It is composed of a monitor 2, a light pen 3, a depth specifying device 4, and a liquid crystal shutter glasses device 5. The concept of the operation of this device will be explained with reference to perspective diagrams in FIGS. 2-3. Note that regarding the specified position, the monitor display surface 2a
In addition to the upper left-right direction (hereinafter referred to as the X direction) and the depth direction, three directions are specified: the upper and lower directions (hereinafter referred to as the y direction). Because of the positioning, the coordinate value in the y direction does not affect calculation of the cursor position for displaying a stereoscopic image, which will be described later.

Therefore, two variables, the X direction and the depth direction, are used as objects of calculation.

First, to explain the perspective relation position between the specified point and the viewpoint, FIG. 2 shows the perspective state of the specified point at the viewpoint height, and in the same figure, point E1? .

The town is the right and left eye positions of observer 10, point E.

is the midpoint between points ER and EL. The straight line E RE t is parallel to the display surface 2a. Furthermore, the midpoint Ec
The intersection of the perpendicular line with the monitor display surface 2a passing through is defined as a point O.

Let the point P be the specified point of the [virtual solid] elephant you are trying to input, and the straight line E. P.EI? Let the intersections of P, E, , P and the monitor display surface 2a be points C, L, and R, respectively. To explain each point set as above, first, monitor surface 2a
The upper point O is the origin of the X-axis, with the horizontal direction of the monitor 2a being the X-axis. Note that point 0 is also the origin of the y-axis, which is perpendicular to the drawing. The point C on the monitor is the midpoint E of the designated point P on the stereoscopic image. This is a point viewed through the monitor surface 2a from a viewpoint (hereinafter referred to as a perspective point), and is inputted to the controller by the uitoben 3 as first input information.

That is, point C is a perspective point of point P, and at the same time is a specified point specified by the observer using the light bend. Note that the value of the X coordinate of the specified point C is assumed to be X. However, when considering the component in the y direction, the X+Y coordinates of the specified point C are (x,
y). The above dots and R represent perspective points on the monitor surface L2a with the left or right eye position E or ER with respect to the specified point P on the CC stereoscopic image as the viewpoint. Then, the position of H on the monitor surface 2a in the perspective view is inputted by the depth specifying device 4 to the controller as second input information, which is the positional information of the depth of the specified point P on the stereoscopic image, and is subjected to calculations described later. It is determined by Eq.

FIG. 3(A) shows a case where the liquid crystal shutter 5a on the left eye side of the liquid crystal shutter glasses device 5 is in a transmitting state, and the liquid crystal shutter 5b on the right eye side is in a blocking state, and furthermore, the monitor surface 2
A cursor for the left eye is displayed on point P as a perspective point based on the viewpoint of the left eye. Therefore, in this state, the observer 10 sees the dotted cursor only with his left eye. On the other hand, the third
In FIG. 3(B), contrary to FIG. 3(A), the liquid crystal shutter 5a on the left eye side is in the shielding state, and the right side is in the transmitting state, and similarly,
A right-eye cursor is displayed on the monitor surface 2a at a perspective point R based on the viewpoint of the right eye at point P.

Therefore, in this state, the cursor at point R is observed only with the right eye.

Then, the above-mentioned dotted and R cursors are displayed alternately on the monitor 2 as images for the left and right eyes in a time-sharing manner.
In conjunction with this display, the liquid crystal shutter glasses 5 are switched between transmitting and blocking, and the observer 10 can observe the designated stereoscopic point P as a stereoscopic image on the monitor 2 using the cursor H. In addition to the dot and R cursor mentioned above,
The lines connecting the specified points that have already been input are also displayed on the monitor 2 at the same time as the cursor as the above left and right eye images.
Observed as a three-dimensional image.

Next, based on the depth information input by the depth specifying device 4, the relative distance Z from the specified point C to R is determined.
and point, R's X coordinate value xL.

The calculation formula for XR will be explained.

Figure 4 shows the viewpoint, designated point C, and monitor surface 2 in Figure 2.
Point E shown in Fig. 4 shows the relationship between the points on a.
Since R, EL, EC, O, P, L, C and R are the same as those in FIG. 2, their explanation will be omitted. When designing the system of this device, the baseline length S is defined as 1/2 of the distance between the left and right eyes (for boat fishing, the baseline length S is set to 3 to 3.5 meters), and Midpoint E of left and right eyes. and the distance g between the origin O of the monitor surface 2a. Set. Then, the coordinates of point C input by Light Ben 3 are X, and the middle point E. Let the distance between and the specified point C be g. Also, from specified point C to specified point P
Let d be the distance to the midpoint E. Let the angle θ be between the origin 0 direction and the designated point P direction.

First, the above distance flail is calculated from the X coordinate value X of point C specified by considering ΔEC0C and the distance 11i1! between the observer 10 and the monitor surface 2a of the system! N
It is expressed using o, and becomes. Also, the relative distance 'aZ is ΔE c P E R
By considering the similarity relationship between ΔCPR and ΔCPR, is obtained, and by substituting equation (1) for g in this equation, is obtained. Note that in equation (1), if the angle of view is small and θ is within 8 degrees, the following can be used as an approximate equation with an error within 1%. On the other hand, it's lit.

The X-axis coordinate value of R is X t, , X R is at point C,
Since it is the midpoint of R, it can be calculated as the 9th order.

xR-xe+Z...(4)xL-x, -Z
(5) It is assumed that the X coordinate is treated with the right direction as + and the left direction as - with respect to the origin O.

Next, to explain the coordinates of points L and R when designated point C includes a component in the y direction, since the direction of both eyes of the system is located in the horizontal direction as described above, for designated point C,
The value in the y direction does not change, and each takes the value of the coordinate value y. On the other hand, regarding the X coordinate, the specified point on the 3D image is projected onto the horizontal section as shown in Figure 2, and the point is treated as point P, and the distances g and d are also converted to the above horizontal section and used in calculations. Then, the above equations (4) and (5) can be used as they are for the X-axis coordinates X R, The coordinates (xy) and dots can be expressed as coordinates (xl,,yo)R″C.

As specific examples of images of lines connecting the cursor positions and cursor points calculated as described above, FIG. 5(^) shows an image for the left eye, and FIG. 5(B) shows an image for the right eye. are alternately displayed on the monitor 2 and viewed through the liquid crystal shutter 5, it can be observed as a three-dimensional image.

The overall configuration of a three-dimensional image processing apparatus showing an embodiment of the present invention based on the above concept is as described with reference to FIG. FIG. 6 is a block diagram of the image processing apparatus of this embodiment. A device having the same functions as the conventional example for the two-dimensional image input device explained with reference to FIG. Explanation of color palette IC% character ldSVRAMles keyboard switch if, SSGlg, and counter 1h will be omitted.

Among other components, the depth specifying device 4 is shown in FIG.
), the second input information regarding the depth position is input to the controller 1 using a knob 4a in a manner described later.

Further, the liquid crystal shutter glasses device 5 is a glasses device having an electronic shutter function that switches between a transmitting state and a shielding state in accordance with the display timing of the right eye image and the left eye image that are alternately displayed on the monitor 2. be.

Further, the CPU 1a takes in the coordinates x, y of the designated point C input by the light ben 3 according to the processing program 1b stored in the ROM, reads the depth information from the depth designation device 4, and calculates the
4) and (5), the coordinates XR and XL are calculated and output to the monitor 2 as images for the left and right eyes.

Furthermore, it is assumed that locus data of a line connecting designated points of the stereoscopic position inputted by the cursor position is stored separately in the VRAMI e for the left and right eyes, and is reproduced on the monitor at the same time as the cursor is displayed.

Next, regarding the input of depth information, as described in the explanation of the concept, when specifying the depth, the distance Md from the monitor surface 2a to the three-dimensional specified point P may be directly substituted into equation (3); In this method, the distance to the object is input in correspondence with a diopter defined as the reciprocal of the distance, which is a unit of optical object distance, so that the input can be more closely matched to the senses of the human eye. To explain the diopter, it was originally a unit used to express the power of a lens, and indicates that it is the reciprocal of the numerical value when the focal length is measured in meters.

It can be converted into a film and similarly used to express the distance to the object point as a reciprocal; when the distance to the object point is D [m], the reciprocal expression is 1/g [diopter].

shall be. (Hereinafter, the value of AP1 distance in units of diopters will be referred to as a "reciprocal distance expression value" or simply "reciprocal distance.") In other words, the depth specifying means operates in proportion to the reciprocal distance. It will be configured.

The effects of doing so will be explained below. Figure 7 shows a model of the human retina,
EL and ER are the centers of the left and right eyes, M is the retina, P' is the target point, and P' is the point projected on the retina by the right eye. Furthermore, S represents 1/2 of the distance between the eyes, h represents the focal length of the eyes, g represents the distance to the target point, and t represents the position of one image on the retina. Distance i! When tIg is expressed as a reciprocal value, it becomes l/g. Then, the reciprocal distance R1 between the image position t and the target point in FIG.
The relational expression with /g is as follows. That is, the position t of the image and the reciprocal distance 1/g of the target point are in a proportional relationship. Therefore, when indicating the distance to a target point, if it is expressed in diopters (units), it will be proportional to the change in the position of the image as seen by the human eye, and in this 3D image processing device, the 3D image is expressed in diopters (units). If the position is input, the input processing will be done in a way that matches human senses. On the other hand, when considering the physical quantities input by the designation device, for example, resistance, length, angle, pressure, or when using a digital counter, the distance to the three-dimensional designation point becomes longer compared to the number of counts. When this happens, it becomes difficult to input distance values directly. Therefore, if the depth information is input at the 111th position described above, this problem will be solved, and furthermore, it will be possible to input depth information that matches the sense of the eye as described above.

The depth device 4 of this embodiment uses a slide volume as shown in FIG. 9(^), and there is a relationship between the set position of the knob 4a and the depth information value inputted thereby in units of diopters. I will explain about it. FIG. 8 shows depth information value D1 corresponding to the slide volume knob position (total resistance value R8).
), and the midpoint EC of l'' shown in Figure 4.
It shows the relationship between the distance g+d from to the three-dimensional designated point P. Note that the above depth information value is the value obtained by subtracting the diopter value of the distance g from the center point EC of the eye to the designated monitor point C minus the diopter value of the distance j7+d to the point P, that is, it is expressed as follows. . Therefore, when the knob 4a is located at the center value D-0, the 3D designated point P is on the monitor surface 2a, and when the value D>0, the point P is deeper than the monitor surface 2a, and when the value D< In the case of O, the point P will be located in front of the monitor surface 2a.

Since the knob 4a is directly related to the diopter (like the upper part), the input operation is performed by an operation that corresponds to the sensation of the observer's mouth.

On the other hand, the resistance value of the slide volume for each knob position is based on 0.5 Ro for the value D-0 knob position, and is determined so that the resistance value changes in proportion to the change in value. Depth information value varies depending on the value of CPU1
is input to the controller 1 via a.

Then, the relative pressure MZ from the specified point C of the points R and L where the cursor is displayed is transformed to the equation (2) and used as an arithmetic expression via the value, Z-DJ)-s is obtained, and further, Substituting equation (1) into g, Z is...
・・・・・・・・・(7) If values are specified by the light pen 3 and the X Cs depth specifying device, the value of the relative distance Z can be calculated.

To explain the operation of the three-dimensional image processing apparatus of this embodiment configured as described above, the observer 10 views the right and left eye images displayed on the monitor 2 through the glasses device 5, thereby realizing the three-dimensional image processing apparatus. While observing the image, specify a desired specified point C on the monitor surface 2a with the light pen 3, and further select a desired value with the knob 4a of the depth specifying device 4.
In the controller 1, the relative distance Z is calculated using the above equation (7). Then, the cursor of the right and left eye images is set to point R (coordinates
) Right and left liquid crystal shutters 5a on top.

It is displayed in accordance with the timing of 5b. And the right,
As left eye image data, the cursor position data or line data connecting the cursors is stored in the VRAM 1e and displayed on the monitor 2 at the same time. Therefore, by moving the specified point C while keeping the light pen 3 in contact with the monitor surface 2a and operating the depth specifying device 4 at the same time, any trajectory including curves can be drawn in real time in the three-dimensional space. can do.

In addition, in Fig. 8 showing the related distance corresponding to the specified depth value indicated by the diopter, the distance to point P is 0 + d.
Regarding the (m) data, in the case of a distance of 1-2 m, the distance corresponding to a value of 0.5 or more is not shown, but this range is an impossible range in reality, assuming a distance of 1 filN + d. is specified by the observer, the controller 1 treats it as if 0.5 has been input as 1iiD.

Further, in the above embodiment, a slide volume was used as the depth specifying device 4;
), or a depth specifying device 41 having an operation knob 41a using a rotary type volume, or a built-in variable resistor as shown in FIG. 9(C),
A depth specifying device 42 having a pedal-type operating section 42a or the like is applicable.

Furthermore, although the above-mentioned depth specifying device was operated independently of the Light Ben 3,
) shows a depth specifying device that is integrated with the above-mentioned light ben, which allows input of the specified point C1 and the depth information value with one hand, and is very easy to use.

The depth designation device 43 shown in FIG. 10(^) is slidably fitted into an outer cylinder 43h in which a designation switch 43c and a sliding contact 43f are mounted, and the inside of the outer cylinder 43h. It is constituted by an inner cylinder 43a. The inner cylinder 43a is equipped with a light receiving element 43d such as a directional SPD at its tip, and inside is a sliding resistance plate whose sliding direction is the longitudinal direction along the opening 43g. 43e is installed. Further, it is assumed that the inner 1f 143a is biased by a spring 43b in the direction in which the light receiving cable 43d is installed. Furthermore, the sliding contact 43f mounted in the outer cylinder 43h is connected to the resistance plate 43 mounted in the inner cylinder 43a.
e, and the resistance value Ra between the terminals changes in proportion to the pushing amount of the inner cylinder 43a.

To explain the operation of the depth specifying device 43 configured as above, first, the outer cylinder is grasped and the light receiving element 43d is pressed onto the monitor surface 2a. When the designation switch 43c is pressed at the designated point C, the coordinates of the designated point C are read into the controller 1. In addition, to input the depth information 11D, press the outer cylinder 43h toward the monitor surface 2a, and press the inner cylinder a.
can be pressed, the resistance value Ra can be changed, and the depth information value corresponding to that value can be input to the controller 1.

Next, the depth designation device 44 shown in FIG. 10(B)
An outer cylinder 44f in which the designation switch 44c is mounted, a pressure-sensitive element 44e whose one pressure-sensitive surface is made of pressure-sensitive ceramic or pressure-sensitive rubber and which is the same as the outer cylinder 44f, and the pressure-sensitive element T.
It is held by the outer tube 44f in contact with the other pressure-sensitive surface of the pressure sensitive surface 44e, and is formed by the light-receiving element 44d having directivity. It is assumed that the light-receiving element 44d is held in the outer cylinder 44f so that when its tip is pressed, the pressing force is transmitted to the pressure-sensitive element 44e.

The operation of this depth designation device 44 is the same as that of the depth designation device 43 described above regarding the input of the designated point C, so the explanation will be omitted. 2a, the amount of force is detected by the pressure sensitive element 44e, and inputted to the controller 1 as the depth information value. Furthermore, this depth designation device 4
4 has a structure in which the pressure-sensitive element 44e is pressed via the light-receiving element 44d, but a structure in which the pressure-sensitive element is exposed to the outside and is directly pressed by the monitor surface 2a may also be used.

Detection elements of the depth specifying device other than those described above include strain gauges, dust cores, SPDs, granular carbon, and the like. Furthermore, as another depth specifying means, FIG. 10(C)
It is also possible to use a digital counter type depth designation device 45 having two buttons 45a and 45b as shown in FIG. This device has buttons 45a and 45
By suppressing b, the corresponding switches are turned on.

OFF, and the depth designation information value is increased or decreased in proportion to the number of pulses. Note that the button 45a side acts to decrease the price, and the button 45b side acts to increase it.

Display points for left and right eye image data in this embodiment,
The calculated coordinates of R (XL, Yo) and (xl?,
yc) and the dot coordinate values actually displayed on the monitor 2 are different, and the conversion formula will be explained. FIG. 11 shows a dot display state diagram of the monitor 2 whose display area is M dots x N dots (M and N are both odd numbers).
The dot coordinates above are shown. Then, dot coordinates (m, n
)n--yc. The coordinates for the right eye are found in the same way.

Next, a modification of the three-dimensional image processing apparatus according to the embodiment of the present invention will be described with reference to FIG. 12.

Among the components shown in FIG. 12, the monitor 52, light ben 53, depth specifying device 54, and liquid crystal shutter glasses device 55 have the same functions as those in the above embodiment, so their explanation will be omitted. The controller 51 has a still video floppy disk device 51e for recording and reproducing image data for left and right eyes input from outside the control unit having the same functions as those of the previous embodiment, or normal two-dimensional analog image data. and recording for the disk device 51e, 11f raw buttons 51a and 51b, eject button 51c, and floppy disk access button group 51.
d, 3D processing and 2D processing selection switch knob 51
f. Palette specification button 51 for specifying the color of the displayed image
It is equipped with g.

In this modification, the input stereoscopic image can be saved on a floppy disk or edited by the disk device 51e, and furthermore, the stereoscopic image taken with a normal still video camera can be used as the original image to perform stereoscopic processing. It has enhanced functionality.

[Effects of the Invention] As described above, the stereoscopic image processing device according to the present invention inputs the center point and depth of a perspective point or a projection point on a display screen by a simple operation regarding input of coordinates of a two-lens stereoscopic image. This device is characterized by the ability to easily perform 3D image processing according to the operator's wishes, and while checking the 3D image in real time, this device sequentially calculates the perspective coordinates of specified points on the 3D image. It has remarkable features such as being able to simultaneously input the perspective coordinates and depth information, and specifying it by fixing one of the perspective point coordinates and depth information and changing the other data. It is suitable for drawing three-dimensional pictures or as an input device for a three-dimensional CAD system.

[Brief explanation of the drawing]

FIG. 1 is a perspective view of a stereoscopic image processing device according to an embodiment of the present invention, and FIG. 2 is a perspective diagram showing the relationship between specified points, perspective points, and viewpoints of the stereoscopic image processing device of the present invention. , Figure 3 (A) , (
B) is a perspective diagram for the left eye and right eye with respect to the above figure 2, Figure 4 is a positional relationship arrangement diagram of the viewpoint, display point, and three-dimensional designation point in the above figure 2, and Figure 5 (^ ), (B) is an image for the left eye in the two-lens stereoscopic image system using the stereoscopic image processing device shown in FIG. 2 above] and an image for the right eye. FIG. 7 is a block diagram of a three-dimensional image processing device; FIG. 7 is a human web model diagram; FIG. 8 is a diagram showing the diopter value and resistance value of the depth designation device of the processing device shown in FIG. 1; (A), (r3), and (c) are the first
External views of each vertical depth specifying device of the processing device shown in the figure, Figures 10 (^), (B), and (C) are the above-mentioned
FIG. 11 is a diagram showing the state of dots displayed on the monitor of the processing device shown in FIG. 1, FIG. FIG. 13 is a perspective view of a conventional image processing apparatus, and FIG. 14 is a block diagram of the conventional apparatus shown in FIG. 13.

Claims (1)

[Claims] (1) Both the left-eye cursor display position and the right-eye cursor display position in binocular stereoscopic image processing are based on the first input information that specifies the in-plane position of the screen and the depth direction position of the same screen. A stereoscopic image processing device characterized in that it can be controlled simultaneously in response to specified second input information.