KR100324392B1 - 3d digital system - Google Patents

Abstract

The present invention relates to a three-dimensional digital input device, and more particularly, to obtain a two-dimensional image signal through an optical system and an imaging unit, and to synthesize the three-dimensional image signal.

The configuration of the present invention for this purpose is largely divided into an image input unit, a stereoscopic image unit, and a screen processing unit.

The image input unit is circularly arranged at 45 degrees with the distance between the horizontal image input unit having 45 degrees of 8 CCDs in order to obtain image information on the side of the object, and the distance between two CCDs of the horizontal image input unit in order to obtain the upper image information of the object. It is divided into three vertical image input units consisting of three CCDs, and the image input unit obtains all necessary 2D image information except the bottom surface of the object by digitizing only two levels of horizontal image input unit and one pair of vertical image input unit. do.

The three-dimensional image unit classifies the two-dimensional image digital signals obtained from the CCDs of the horizontal image input units into phase signals and color signals, and uses the phase signals and color signals obtained and classified from two CCDs arranged in parallel from the reference CCD to the eighth CCD. First, three-dimensional horizontal center. The data of the upper part is three-dimensionalized from the horizontal center line and the lower part is three-dimensionally compared with the position data of the three-dimensionalized horizontal center part.

The screen processor processes the filed stereoscopic image data and displays the stereoscopic image on the screen according to the direction and size desired by the user.

Description

3D digital input device {3D DIGITAL SYSTEM}

The present invention relates to a three-dimensional stereoscopic image. More specifically, the present invention relates to a three-dimensional digital input unit that obtains two-dimensional information on all objects such as a person or a product and three-dimensionally converts the image into a three-dimensional image.

The fields related to 3D stereoscopic image are largely divided into distance image acquisition, feature extraction, and recognition. Among these fields, a direct relation is the binocular stereoscopic method.

In the range of distance image acquisition, laser light passes through the lens to create a stripe, the camera receives this information, calculates the distance value, obtains the range image, and moves the focal length of the camera to move the distance image to a clear object. The binocular system obtains the distance information of one object in two images using the active focusing method, the focus / nonfocus method for obtaining distance image by quantitatively obtaining the degree of blurring, and the principle of identifying the object position from the angle difference between human eyes. Stereoscopic methods have been researched and developed.

The binocular stereoscopic method was mainly applied to stereoscopic digital cameras and the like to image only a part of an object. Therefore, the three-dimensional shapes in the virtual space were processed in the real world by graphic processing or the whole graphic work, which has the limitation of reducing the realism in the virtual space.

In order to solve these drawbacks, real-time stereoscopic images having the size and color of real objects can be made and dataized in a short time so that they can be displayed or tested alone or in combination of stereoscopic images of several objects in virtual space.

The present invention digitizes the image information in an analog state from an image input unit comprising an optical system and a CCD image pickup device (hereinafter, an image pickup unit), and three-dimensional image unit three-dimensionally to have the same dimensions and color information as the real object. Next, a three-dimensional digital input device for creating a stereoscopic image file through a separate correction operation and displaying it on the screen according to the direction and angle desired by the stereoscopic image processing unit.

1 is a principle of obtaining three-dimensional stereoscopic image position information of the center line according to an embodiment of the present invention

2 is a principle of obtaining position information of a 3D stereoscopic image excluding a centerline according to an exemplary embodiment of the present invention.

Figure 3 is a basic layout of the three-dimensional digital input device image input unit according to an embodiment of the present invention

4 is a block diagram of a three-dimensional digital input device according to an embodiment of the present invention;

5 is a movable and rotating CCD of a prefabricated three-dimensional digital input device image input unit according to an embodiment of the present invention;

<Description of the symbols for the main parts of the drawings>

1,2,26: optical system 3,4,27: imaging unit

5-21,32,33,34 CCD 22,23,24,25 Lamp

35: stepping motor

The structure of the present invention for achieving the above object is largely divided into an image input unit, a stereoscopic image unit, and a screen processing unit.

The image input unit is circularly arranged at 45 degrees with the distance between the horizontal image input unit having 45 degrees of 8 CCDs in order to obtain image information on the side of the object, and the distance between two CCDs of the horizontal image input unit in order to obtain the upper image information of the object. It is divided into three vertical image input units consisting of three CCDs, and the image input unit obtains all necessary 2D image information except the bottom surface of the object by digitizing only two levels of horizontal image input unit and one pair of vertical image input unit. do.

However, the image input unit is not necessarily composed of two sets of horizontal image input units and one set of vertical image input units, and either one horizontal image input unit or one vertical image input unit is selectively used during production and operation for obtaining two-dimensional image information according to the size of the object. May be excluded. This means that the size of the image information of the side surface of the object is limited according to the size of the object and can be adjusted accordingly. In addition, the image input unit may be manufactured in a fixed space in a limited space, or may be manufactured in an adjustable type that can adjust the diameter of the CCD circular arrangement according to the size of the object.

The three-dimensional image unit classifies the two-dimensional image digital signals obtained from the CCDs of the horizontal image input units into phase signals and color signals, and uses the phase signals and color signals obtained and classified from two CCDs arranged in parallel from the reference CCD to the eighth CCD. First, three-dimensional horizontal center. The data of the upper part is three-dimensionalized from the horizontal center line and the lower part is three-dimensionally compared with the position data of the three-dimensionalized horizontal center part.

The stereoscopic image data of one set of three-dimensional horizontal image input units is merged with data of another set of horizontal image input units at the middle of each horizontal center line. Here, if the image signals obtained from the three CCDs of the vertical image input unit are processed in the same manner, a realistic stereoscopic image is produced. The completed stereoscopic image is saved as a file.

The screen processor processes the filed stereoscopic image data and displays the stereoscopic image on the screen according to the direction and size desired by the user.

The above configuration will be described in detail with reference to the accompanying drawings.

1 and 2 are illustrated to explain the principle of obtaining position information of a 3D stereoscopic image. The reason for separating and obtaining the position information of the stereoscopic image is to overcome the principle that the clock distortion of the two-dimensional image signal obtained from the image pickup unit, that is, all the optical systems increases as the distance from the center becomes large To do this, we created an algorithm that processes the data into the core and other areas.

1 is a diagram showing a principle of obtaining three-dimensional stereoscopic image position information of a center line. The reference optical system 1 and the imaging unit 3 are referred to as the reference CCD 5, the left adjacent optical system 2, and the imaging unit 4. Are the same CCDs as those in Fig. 2A. This is that Fig. 1 and Fig. 2 use image signals obtained by CCDs installed at the same position with respect to the same object, and only the positions of points B and C are different.

The reference CCD 5 and the adjacent CCD 6 are positioned at the same distance from the same point A, and arranged so that the CCD element centers coincide. At this time, the angle formed by the reference CCD 5 and the adjacent CCD 6 is 45 degrees, but if the change of the shape of the object is simple, the angle may be arranged at 60 degrees. The arrangement of the CCD, or optical system, is largely different from that of the prior art in which the optical systems are arranged at intervals corresponding to the left and right eye widths of a person, and different angles are used. In the conventional method, since the information obtained from the two optical systems is similar, Distortion is similar in the field of clock distortion between the two-dimensional image information obtained in the negligible or correctable error range and the actual shape of the object in the same spatial plan, so it is easy to obtain stereoscopic images but obtain relatively accurate position information. There is a disadvantage that is not easy. On the other hand, the input devices of the present system have a large arrangement angle of the CCD, so that the similarity of the obtained information is greatly reduced and the aspect of distortion is very different, but relatively accurate position information can be obtained.

For this purpose, the device center is not a physical center, but a center obtained by shifting the focal length of the acquired image data, and each CCD is disposed at the same distance and the same angle as the center of the designated horizontal plan in the input device.

When the reference CCD 5 and the adjacent CCD 6 arranged as shown in FIG. 1A form the point B of the object as a two-dimensional digital image signal and are scattered on the screen, FIGS. 1B and 1C are shown. In the generated digital image signal, a phase signal and a color signal are used to distinguish the areas on the same horizontal line, and when the same area and the same point are found, the point B becomes. As shown in FIGS. 1B and 1C, when the uv axis is set around the device center, the coordinate of the point B becomes (u ₁ , 0) in FIG. 1B and (u ₂ , 0) in FIG. 1C. Here, the v axis is the same as the y axis in the xyz three-dimensional coordinate system, and the u axis is defined by the binary linear system of x and z.

Using this principle to calculate the three-dimensional coordinates of point B (x ₁ , y ₁ , z ₁ )

Becomes

By using the above equation, it is possible to synthesize each two-dimensional image signal in the region which is calculated in advance into a three-dimensional stereoscopic image signal. This three-dimensional video signal has position information up to one decimal place. In addition, the area inside the area is visible in the basic CCD (5), but in the adjacent CCD (6) to three-dimensional synthesis of the part is hidden by using a CCD positioned at -45 degrees from the basic CCD (5) three-dimensional Or three-dimensional synthesis by finding the same point using contour and phase signals from three CCDs including two adjacent CCDs located at ± 45 degrees with the basic CCD.

The synthesized stereoscopic image signal is two-dimensionalized by inversion and compared with the two-dimensional image signal as input information. If there is an error, the blurring according to the focal length is corrected in the 2D video signal by using the position information of the 3D video signal, and then the error regions are synthesized again into the 3D stereoscopic video signal and compared again. Repeat until the error is within the required range.

2 is a diagram illustrating a principle of obtaining position information of a 3D stereoscopic image excluding a center line. The arrangement of the reference CCD 5 and the adjacent CCD 6 is the same as that of FIG.

Synthesizing the positional information and the color information of the three-dimensional video signal synthesized according to the principle of FIG. 1 takes time. Therefore, in the present invention, the three-dimensional synthesis of the remaining regions excluding the central portion (a plan of y = 0) estimates each of the x and z coordinates in the y = 1 ± 0.5 region based on the position information of the central portion synthesized as described in FIG. can do. Also, the points on the horizontal line of the two-dimensional video signal v = + 1 obtained by the reference CCD 5 and the adjacent CCD 6 are present within the range of y = 1 ± α rather than y = 1 due to clock distortion. α increases as y increases.

The principle of obtaining location information of 3D stereoscopic image excluding center line is as follows.

When the reference CCD 5 and the adjacent CCD 6 arranged as shown in FIG. 2A make the point C of the object into a two-dimensional digital image signal, and are scattered on the screen, FIGS. 2B and 2C are shown. The coordinate value of the point C ₁ obtained from the reference CCD 5 becomes (u ₁ , v ₁ ) and the coordinate value of the point C ₂ obtained from the adjacent CCD 6 becomes (u ₂ , v ₂ ). Point C in Fig. 2A is point C ₁ in Fig. 2B and point C ₂ in Fig. 2C.

First, in order to estimate z as the average of the presynthesized surrounding points, let z _{l be} the left point, z _{r the} right point, and z _b as the bottom point.

Using the estimated z and the coordinates of the reference point C ₁ , x, y

Here, f means the distance from the origin in space to the element center of the imaging surface.

Calculate the estimated coordinate value of point C ₂ using the obtained xyz coordinate value

Becomes Compensate for the blurring of the image signal of point C ₁ according to the focal length within the area of ± 1 on the uv axis around the estimated coordinates of point C ₂ , and find the point where the phase signal and the color signal are the same or within a certain error range. Change the value of the point to the coordinate value of point C ₂ as uv value. Using the coordinates of the reference point C ₁ and the changed point C ₂

Becomes If this z value is within the required range compared with the z values of the combined peripheral points, use this to recalculate Equation (7) (8) (9) (10) (11) to find the z value and Compare with z and repeat until the difference is ± 0.1. If the z value is not within the required range, the calculated information is temporarily stored, the next point is synthesized, the positional relationship is compared with the adjacent points, and the equation (7) (8) (9) (10) (11) is calculated again. Determine the location of three points next to each other. The points used in the above process by interrelationships may be three or more points.

FIG. 3A is a basic layout diagram of a horizontal image input unit of a 3D digital input unit for synthesizing a side portion of an object into a stereoscopic image, but may include an upper portion thereof. As shown in Fig. 3A, eight CCD (7) (8) (9) (10) (11) (12) and (13) (14) are arranged at 45-degree intervals from the center point, respectively, so that eight two-dimensional images are provided. Get a signal. The synthesized 2D video signal is synthesized for each line with ± 0.5 range of Y-axis unit distance as 1 line, and the synthesis sequence is as follows.

Start by synthesizing (7) CCD as the reference CCD on the right side (8) as the adjacent CCD, and comparing and synthesizing a total of 8 times per line using two CCDs adjacent to each other in the counterclockwise direction. 7) Synthesis of 1 line by using CCD (14) on the left as the adjacent CCD as the reference CCD. The reference CCD is (7) (9) (11) (13) located on the x, z axis, and the adjacent CCD is (8) (10) (12) (14). Synthesize using adjacent CCDs. When synthesizing according to the above arrangement and order, x ₁ and z ₁ in Eq. (1) (3) are changed or +,-in Eq. (3) are changed depending on which axis the reference CCD is on. based on the CCD to obtain the value z ₁ x ₁ to the value unit of length is in the z-axis, is because to obtain the value x ₁ to a value z ₁ in a unit length is in the x-axis.

3B is a basic layout diagram of a vertical image input unit of a three-dimensional digital input device. Three CCDs 17, 18, 19 and a CCD 15 representing two sets of horizontal image input units for synthesizing an upper part of an object into a stereoscopic image are shown in FIG. (21), (16) and (20) and the upper and lower illuminations.

In synthesizing the upper three-dimensional image, after the three-dimensional three-dimensional image signal of the side part is synthesized by the horizontal image input unit, the CCD (20) is used as the reference CCD and the CCD (19) on the right is the adjacent CCD. Start synthesis, and compare and synthesize a total of four times per line using two counterclockwise adjacent CCDs, and use (16) CCD as the reference CCD and (17) on the left side as the adjacent CCD. Finish the synthesis of one line.

In addition, the illumination 16, 20, 15, 21 is provided with two sets of eight each on the top and bottom of the CCD used in the horizontal image input unit. The angle of illumination of the object is determined between 35 degrees and 55 degrees according to the distance between the center distance and the horizontal image input unit.

4 is a block diagram of a three-dimensional digital input device. The optical system 26 and the image pickup unit 27 digitize the image signal captured here as an image input unit by correcting noise in the signal processing unit 28. FIG. The digitized two-dimensional image signal is synthesized into a three-dimensional stereoscopic image by the stereoscopic image unit 29 and converted into a two-dimensional image signal by the screen processor 30 and displayed on the monitor 31.

5 is a diagram illustrating a movable and rotating CCD of an image input unit of a prefabricated three-dimensional digital input unit that can be moved and adjusted according to another embodiment of the present invention. FIG. 5A illustrates a rack and pinion using a DC motor on a vertical frame. 5b shows the movement of the center distance of the circularly arranged CCD. The connecting frame is connected to the movable center lower frame, and the vertical frame is assembled to the center of the image input unit. The distance f can be made large.

5C is a diagram showing the rotation of the CCD 34 using the stepping motor 35.

According to the embodiment of the present invention as described above, by realizing the whole of the object three-dimensional display using the information of the three-dimensional object in the cyber space, it is possible to ensure the authenticity of the product on the Internet, By making comparisons and coordination possible, customers can have the opportunity to experience virtual experiences, which can increase the rationality of product selection and increase the satisfaction of indirect shopping through the Internet. Seen to be.

Claims (8)

Optical systems circularly arranged at an angle of 45 degrees with respect to the object to form an image on the object; An imaging unit which outputs an electrical signal corresponding to each subject formed in each of the optical systems; An image input unit including two horizontal image input units in which eight fixed CCDs each including an optical system and an imaging unit, that is, CCD imaging elements, are circularly arranged on a horizontal plan, and one vertical image input unit circularly arranged on three fixed CCD vertical planes; A signal processor which processes an electrical signal corresponding to each subject into a digital two-dimensional image signal; A stereoscopic image unit for dividing each digital two-dimensional image signal into a reference CCD and a neighboring CCD to synthesize a three-dimensional stereoscopic image signal; And a screen processing unit converting the synthesized 3D stereoscopic image signal into a 2D image signal having a desired size and direction and displaying the same on a monitor. The method of claim 1, 3D digital input device, characterized in that each CCD is circularly arranged at a 45 degree angle. The method of claim 1, And processing the electrical signals output from the respective imaging units as digital video signals, storing them in a memory, loading the necessary digital video signals, and synthesizing them into three-dimensional stereoscopic video signals. The principle of claim 1 is the same, The image input unit is composed of only two sets of horizontal image input units, the lower one set of horizontal image input unit CCD is fixed type, the upper one set of horizontal image input unit CCD is adjustable three-dimensional digital input device that can be adjusted vertically. The principle of claim 1 is the same, The image input section is composed of only two sets of horizontal image input sections. Each CCD is rotated by ± 45 degrees in the vertical direction. Two CCDs located on the same vertical line can move up and down on the same vertical frame. Prefabricated 3D digital input device that connects the connecting frame to the movable central subframe connected by 45 degrees and 8 orientations and assembles the vertical frame to adjust the center distance of the circular frame. The method of claim 5, Three-dimensional digital inputs that each CCD can rotate and move up and down on the same frame. The method of claim 5, Three-dimensional digital inputs where each vertical frame is assembled to a subframe. The method of claim 5, Three-dimensional digital inputs where each connecting frame is assembled to a movable central subframe.