Cylindrical 3D Reconstruction via Radius-Axis Bundle Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial endoscopes face challenges in accurately determining the size and location of defects within pipes due to the limitations of incorporating measurement and pose detection components, and existing methods like SfM and V-SLAM struggle with scale-drifting errors during real-time three-dimensional reconstruction of cylindrical shapes.
Innovation Solution
An image processing system that estimates the camera's position and pose by extracting feature points from images, reconstructing them in three-dimensional space, and performing bundle adjustment to minimize differences between the cylindrical shape's radius and distances from the center axis, thereby correcting camera positions and coordinates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If SfM or V-SLAM methods are used for real-time three-dimensional reconstruction, then inspection efficiency is improved, but scale-drifting errors occur during reconstruction
Solution Approach 1:
The patent implements a feedback mechanism by calculating the distance from each three-dimensional point to the estimated center axis and comparing it with the expected cylindrical radius. The bundle adjustment process uses this feedback to iteratively correct camera positions and refine the center axis estimation, thereby eliminating scale-drifting errors while maintaining real-time reconstruction capability
Solution Approach 2:
The patent changes the parameter optimization approach by incorporating the cylindrical constraint (distance from center axis should equal radius) as an additional optimization condition in the bundle adjustment process. This modifies the standard SfM/V-SLAM parameters to include center axis position and cylindrical geometry constraints, enabling accurate reconstruction without scale-drifting
2Adaptability or versatility
If the probe diameter is kept small (several millimeters) for inspection, then adaptability to pipe interiors is improved, but mounting area for measurement components is reduced
Solution Approach 1:
The patent replaces mechanical measurement components (such as physical calipers or contact sensors) with an optical-based three-dimensional reconstruction system. By using the camera to capture images and computationally reconstruct the pipe interior geometry, the system eliminates the need for additional mechanical measurement devices, keeping the probe diameter small while maintaining measurement capability
Solution Approach 2:
The patent creates a digital copy (three-dimensional model) of the pipe interior geometry through image processing and reconstruction algorithms. This virtual model serves as a measurement reference, replacing the need for physical measurement components and enabling accurate defect size determination without adding mechanical parts to the probe
Data Source
AI summary
An image processing system, which estimates a position and a pose of a camera and performs three-dimensional reconstruction processing of an object having a cylindrical shape by using a photographed image acquired by photographing an inside of the object with the camera, includes a processor. The processor estimates a first center axis of the cylindrical shape by using a group of three-dimensional points obtained by reconstructing, in a three-dimensional space, a group of feature points extracted from the photographed image. In addition, the processor performs bundle adjustment for correcting the position and the pose of the camera and the coordinates of the group of three-dimensional points by using a condition for minimizing a total sum of differences between a radius of the cylindrical shape and respective distances from the first center axis to individual three-dimensional points that constitute the group of three-dimensional points.