Global Positioning from a Single Image of a Rectangle in Conical Perspective

Area of Science:

• Computer Vision
• Robotics
• Geometric Modeling

Background:

• Accurate camera pose estimation is crucial for 3D reconstruction and robotic applications.
• Traditional methods often require multiple views or known scene features.
• A method utilizing basic descriptive geometry principles for camera calibration is needed.

Purpose of the Study:

• To present a method for determining camera positioning and orientation from an image of a known rectangle.
• To leverage descriptive geometry for 3D pose estimation.
• To validate the proposed method experimentally.

Main Methods:

• Applying three turns and one translation to obtain dihedral projections of a reference rectangle.
• Projecting the rotated rectangle onto an elevation plane from a specific viewpoint.
• Implementing inverse perspective transformation to recover 3D coordinates from the 2D image.
• Experimental validation using camera imaging and known rectangle dimensions.

Main Results:

• Successfully obtained dihedral projections and conical perspective images of the rectangle.
• Developed and applied an inverse perspective transformation method.
• Experimentally verified the camera's 3D positioning and orientation using image data.

Conclusions:

• The proposed method accurately determines camera pose using basic descriptive geometry and a single image of a known rectangle.
• This technique offers a viable solution for camera calibration and 3D scene understanding.
• The experimental validation confirms the practical applicability of the method.