KR19980085987A - Distance Information Acquisition Method in 3D Shape Restoration System - Google Patents

Abstract

The present invention relates to a distance information obtaining method for obtaining distance information of an object in consideration of the scale of the object present between two-dimensional distance images obtained in a three-dimensional shape reconstruction system. To this end, the method comprises the steps of: obtaining at least two or more distance images of a predetermined object in various directions; Extracting a scale of a predetermined object existing in each distance image acquired in the obtaining step; Extracting a scaling factor of each distance image with respect to the reference distance image by using the scale extracted in the scale extraction step; Re-forming corresponding distance images having the same scale of a predetermined object present in each distance image by using the scaling coefficient extracted in the scaling factor extraction step; And obtaining distance information of the target object existing in the formed distance images, respectively. Therefore, there is an advantage that can obtain accurate distance information for each distance image.

Description

Distance Information Acquisition Method in 3D Shape Restoration System

The present invention relates to a method for acquiring range data in a three-dimensional shape reconstruction system, and in particular, distance information for obtaining accurate distance information in consideration of scales between objects present in the acquired distance image. It relates to the acquisition method.

The 3D shape reconstruction system is one of the active researches in the field of computer vision, and the distance information (or the distance information of the target object present in the 2D range image of the target object obtained from various directions) 3D shape of the target object is processed through registration process to obtain transformation parameters and correspondence points between distance images using integration data and integration process that synthesizes a complete model. To restore. When reconstructing the three-dimensional shape, the distance information on the object represented by the two-dimensional distance image must be accurately obtained so that subsequent registration or integration processing can be efficiently performed and the correct three-dimensional shape can be restored. do.

However, in the currently proposed 3D shape restoration system, a plurality of distance images obtained by shooting a target object in various directions while moving the camera or by fixing the camera and manually moving the target object to obtain distance images in various directions are obtained. The case where the scale of the object present in the different may occur.

For example, as shown in FIG. 1A, the distance image obtained when the camera 10 photographs the object at a point A away from the object 20 is as shown in FIG. 1B. When the distance image obtained when the camera 10 photographs the object at the point B away from the distance 20) is the same as that shown in FIG. 1C, as shown in FIGS. 1B and 1C, the shooting distance is relatively far. The scale of the object in the distance image acquired at the point is smaller than the scale of the object in the distance image acquired at the point A which is relatively close. The difference in scale with respect to the object between the acquired distance images may be caused by an incorrect operation on the shooting distance of the camera even in a controlled state to be photographed at the same distance mechanically.

However, in the existing three-dimensional shape restoration system, since the scale of the target object existing in the acquired distance image is of course the same, the scale of the target object existing in the obtained distance image is different from each other. If the obtained distance information is used as it is, not only an incorrect registration process is performed but also a difficulty in restoring a perfect three-dimensional shape.

Therefore, the present invention has been devised to improve the above-described drawbacks, and the distance image acquired in consideration of the scale of the target object existing between the two-dimensional distance images obtained in the three-dimensional shape reconstruction system is used in the distance image. It is an object of the present invention to provide a distance information obtaining method for obtaining distance information of an object.

The distance information acquisition method according to the present invention for achieving the above object, in the system for reconstructing a three-dimensional shape for a predetermined object, obtaining at least two or more distance images of the predetermined object in various directions step; Extracting a scale of a predetermined object existing in each distance image acquired in the obtaining step; Extracting a scaling factor of each distance image with respect to the reference distance image by using the scale extracted in the scale extraction step; Re-forming corresponding distance images having the same scale of a predetermined object present in each distance image by using the scaling coefficient extracted in the scaling factor extraction step; And obtaining distance information of the target object existing in the formed distance images, respectively.

1 is an exemplary diagram between a photographing distance and a acquired distance image in a conventional three-dimensional shape restoration system;

2 is a schematic block diagram of a three-dimensional shape restoration system for performing a method for obtaining distance information according to the present invention;

3 is a flowchart according to an embodiment of a method for obtaining distance information according to the present invention;

4 is a flowchart of another embodiment of a method for obtaining distance information according to the present invention;

Explanation of symbols for main parts of the drawings

10: camera 20: object

200: processor 220: grabber

210: projector

The above and other objects and various advantages of the present invention will become more apparent from the preferred embodiments of the invention described below with reference to the accompanying drawings by those skilled in the art.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Figure 2 is a schematic block diagram of a three-dimensional shape restoration system for performing the distance information acquisition method according to the present invention, the processor 200, the overall control so that the distance information can be obtained according to the present invention, three-dimensional shape The processor 200 may control the target object 20 for restoring the image, the camera 10 controlled by the processor 200, and the image captured by the camera 10. It is composed of a grabber (Grabber 220) for converting and transmitting the data to the data and the projector 210 is controlled by the processor 200 to emit light to the target object (20).

3 is a flowchart illustrating an embodiment of a method for obtaining distance information according to the present invention.

Next, the method for obtaining distance information according to the flowchart shown in FIG. 3 will be described in detail with reference to FIG. 2.

First, the processor 200 generates a plurality of distance images of various directions of the object object 20 through the camera 10 and the grabber 220 by using one of several methods conventionally presented in step 301. Acquire.

When a plurality of distance images obtained through the grabber 220 are transmitted, the processor 200 proceeds to step 302 and performs sampling processing in the same number of x and y axis directions for each of the acquired distance images. That is, as shown in FIGS. 1B and 1C, the sampling process is performed in the x-axis direction and the sampling process is also performed in the y-axis direction irrespective of the scale of the object image present in each distance image. At this time, since the sampling process is performed in the same manner as the sampling method used when the existing distance information is acquired, the number of data in the x-axis direction and the number of data in the y-axis direction are sampled in each distance image.

Next, the processor 200 proceeds to step 303 and calculates the scale ratio between each distance image using the number of samplings in the x or y-axis direction of the sampling data of each distance image, that is, the same number from each distance image. Since the sampling is performed, when the scale of the target object existing in the distance image is large as shown in FIG. 1B, the number of sampling data will be large, and when the scale of the target object existing in the distance image is small as shown in FIG. 1C. Since the number of sampling data is small, the scaling ratio between the distance images is extracted, the scaling factor between the distance images is extracted, and the flow proceeds to step 304. Here, the scaling factor is used as information for controlling the enlargement or reduction ratio of the object present in the distance image.

In step 304, the processor 200 reconstructs a distance image corresponding to the acquired distance images by using the scaling factor of each extracted distance image. That is, the distance image is enlarged or reduced or the distance image is obtained as it is obtained according to the scaling factor. In this case, the distance image is obtained as it is when the scaling factor is 1.

As described above, when the distance image having the same scale of the object present in each distance image is formed again, the processor 200 proceeds to step 305 and provides distance information on the object present in the formed distance image as in the conventional method. Is obtained and returned.

4 is a flowchart of another embodiment of a method for obtaining distance information according to the present invention, which will be described below with reference to FIG. 2.

First, in step 401, the processor 200 acquires a plurality of distance images as in step 301 of FIG. In operation 402, the bounding box size of the target object existing in each of the obtained distance images is extracted. That is, a bounding box is set on the object for each distance image, and the size of the bounding box is extracted.

In operation 403, a scaling factor of each distance image is extracted using a ratio of bounding boxes of the extracted distance images. That is, the scaling factor for each distance image is extracted by detecting the ratio of the size of the bounding box of the other distance image based on the size of the bounding box of the reference distance image.

The processor 200 proceeds to step 404 to form a distance image in which the distance image is reduced or enlarged or obtained using the scaling factor of each extracted distance image.

As described above, when a distance image having the same scale of the object present in each distance image is formed, the processor 200 proceeds to step 405 in which the distance information of the object present in each distance image is obtained. Get and return

As described above, in the 3D shape reconstruction system, by forming a distance image in which the scales of the target objects present in the plurality of distance images obtained by using the scale coefficients are obtained to obtain corresponding distance information, Even if the scales present in the distance image acquired due to the error of the shooting distance are different, accurate distance information about the object can be obtained, so that an efficient registration process can be expected and a more perfect three-dimensional shape can be restored.

Although the present invention has been described as the above-described embodiment, those skilled in the art will appreciate that various changes and modifications can be made without departing from the spirit of the present invention. Therefore, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification but should be defined by the claims.

Claims (3)

In a system for restoring a three-dimensional shape for a predetermined object, Obtaining at least two distance images of the predetermined object in various directions; Extracting a scale of the predetermined target object existing in each distance image acquired in the obtaining step; Extracting a scaling factor of each distance image with respect to a reference distance image by using the scale extracted in the scale extraction step; Re-forming corresponding distance images having the same scale of the target object present in each distance image by using the scaling coefficient extracted in the scaling factor extraction step; And obtaining distance information of the target object existing in the re-formed distance images, respectively. In a system for restoring a three-dimensional shape for a predetermined object, Obtaining at least two distance images of the predetermined object in various directions; Performing a two-dimensional sampling process for the predetermined target object existing in each distance image acquired in the acquiring step; Extracting sampling interval data from each distance image by using sampling data obtained in the sampling processing step; Extracting a scaling factor for each distance image by extracting a ratio of a sampling interval of each distance image to a reference distance image by using the extracted sampling interval data; Using the extracted scaling coefficients, re-forming corresponding distance images having the same scale of the target object present in each distance image; And obtaining distance information of the target object existing in the re-formed distance images, respectively. In a system for restoring a three-dimensional shape for a predetermined object, Obtaining at least two distance images of the predetermined object in various directions; Extracting a bounding box size for the predetermined target object existing in each distance image acquired in the acquiring step; Detecting a ratio of the bounding box size of each distance image to the reference distance image using the extracted bounding box size of each distance image; Extracting scaling factors corresponding to each distance image by using a ratio of bounding box sizes of the distance images; Using the extracted scaling coefficients, re-forming corresponding distance images having the same scale of the target object present in each distance image; And obtaining distance information of the target object existing in the re-formed distance images, respectively.