KR-102963988-B1 - Methods and devices for estimating camera coordinate system of augmented-reality device using hybrid marker pattern

Abstract

The present invention relates to a method for estimating a camera coordinate system using image recognition technology and composite markers. The camera coordinate system estimation method comprises: (a) a step of designing a composite marker pattern and manufacturing an actual composite marker pattern tool to determine the setting of a world coordinate system, the setting of an element marker pattern, and the setting of an element marker geometric transformation, in order to align the coordinate systems of the virtual space and the real space of an AR device to be set; (b) a step of determining a central element marker and a surrounding element marker for stable calculation regarding the camera viewpoint of a given AR device; (c) a step of estimating a camera coordinate system relative to a central element marker by applying an interpolation method to the central element marker and the surrounding element markers; and (d) a step of converting the camera coordinate system relative to the central element marker into a coordinate system relative to the world coordinate system. The physical equipment is an equipment having a composite marker pattern on one surface, and it is preferable that the composite marker pattern is composed of a composite marker including a plurality of element markers.

Inventors

• 임인성
• 최규진
• 안재풍

Assignees

• 서강대학교산학협력단

Dates

Publication Date

20260511

Application Date

20191224

Claims (5)

1. (a) A step of placing a composite marker device for coordinate system estimation, having a composite marker pattern on one surface (wherein the composite marker pattern is characterized by being composed of a composite marker comprising a plurality of element markers on a single plane), in order to align the coordinate systems of the virtual space and the real space of the AR device to be set, and then acquiring an image including the composite marker pattern; (b) a step of determining a central element marker and a peripheral element marker for stable calculation with respect to the camera view of a given AR device from the composite marker pattern of the image obtained above; (c) a step of estimating the camera coordinate system relative to the center element marker by applying interpolation to the center element marker and the surrounding element markers; and (d) A step of converting the camera coordinate system relative to the center element marker to a coordinate system relative to the world coordinate system; A coordinate system estimation method utilizing image recognition technology equipped with composite markers.
2. A method for estimating a coordinate system using image recognition technology and composite markers, characterized in that, in claim 1, the coordinate system correction method using the image recognition technology and composite markers is applied to an AR device.
3. In claim 1, the physical equipment of step (a) is an image recognition technology characterized by coordinate system estimation in AR space and a coordinate system estimation method utilizing composite markers.
4. delete
5. A composite marker pattern is provided on one side surface, and A composite marker device for estimating a coordinate system, characterized in that the above composite marker pattern is composed of a composite marker including a plurality of element markers in a single plane.

Description

Method and devices for estimating camera coordinate system of augmented-reality device using hybrid marker pattern The present invention relates to augmented reality (AR) technology, and more specifically, to a method for stably estimating the position and orientation of a dynamic augmented reality (AR) device in actual real space, that is, a camera coordinate system (hereinafter camera coordinate system) based on a camera frame, and to an apparatus necessary for the same. Augmented reality (AR) technology is a key technology that enables users to connect real and virtual spaces for direct interaction. The blending of these two spaces allows virtual objects to be projected into the real world, thereby providing a realistic and free experience. However, conventional AR technologies often produce unstable results when users move the device; if the camera frame coordinate system is estimated by recognizing a single fixed marker—as in traditional methods—it frequently leads to image processing errors and noise that may occur during marker recognition. Since the coordinates of the real and virtual spaces do not exactly align, AR technologies cause users to feel a sense of unfamiliarity due to the misaligned coordinate systems. To mitigate this, it has been necessary to perform a coordinate system re-recognition process or impose restrictions on the range of movement. However, if the AR device's camera coordinate system can be recognized in the real world in a way that is robust against errors, the sense of unfamiliarity in AR can be significantly reduced. Consequently, the scope of application and effectiveness of AR are expected to be substantial. FIG. 1 is a flowchart sequentially illustrating a method for estimating a camera coordinate system according to a preferred embodiment of the present invention. FIG. 2 is an example of a composite marker pattern used for estimating the coordinate system of an AR device in a camera coordinate system estimation method according to a preferred embodiment of the present invention. FIG. 3 is a figure illustrating a method for finding the element marker that most matches the camera's line of sight when setting the coordinate system of an AR device in a camera coordinate system estimation method according to a preferred embodiment of the present invention. FIG. 4 illustrates the overall process of the camera coordinate system estimation method according to a preferred embodiment of the present invention. It shows a method of finding multiple element markers that are most closely aligned with the camera's line of sight when setting the coordinate system of an AR device, and correcting using the coordinate system estimated from the element markers. Figure 5 shows the coordinate system estimation result in which only composite markers are recognized when the present invention is not applied to the camera of an AR device. Figure 6 shows the result of correcting the coordinate system estimation error of a composite marker using element markers when the present invention is applied to the camera of an AR device. The present invention aims to stably and robustly estimate the relative geometric transformation between the coordinate system of the camera mounted on the AR device and a designated coordinate system in real space when a user moves the AR device (i.e., to estimate the camera coordinate system in real space). Its main features are summarized as follows. First, the world coordinate system in real space is set at the center point of the composite marker pattern presented in the present invention in a manner traditionally used in marker recognition methods. Second, for each element marker constituting the composite marker pattern, an element marker coordinate system is defined at its own center point. Third, the geometric transformation of each element marker is known in advance. Therefore, when the camera viewpoint of the current frame is determined, if the camera coordinate system relative to a specific element marker is estimated using the existing marker recognition method, the camera coordinate system based on the world coordinate system can be easily estimated using the element marker geometric transformation. Fourth, for a given camera viewpoint, after determining a central element marker capable of stable image processing, a method is adopted to estimate the camera frame relative to the central element marker using a composite method by additionally using one or more surrounding element markers if possible depending on the situation. Fifth, we present tools that can effectively carry out this process. Hereinafter, a method and apparatus for estimating a camera coordinate system of an augmented reality device using a composite marker pattern according to a preferred embodiment of the present invention will be described in detail with reference to the attached drawings. FIG. 1 is a flowchart sequentially illustrating a method for estimating a camera coordinate system ac