3D Reconstruction Using Nominal Camera Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D reconstruction methods face challenges in achieving high accuracy when camera intrinsic parameters are not calibrated, leading to low reconstruction accuracy and optimization divergence.
Innovation Solution
A 3D reconstruction method that uses nominal intrinsic parameters and initial distortion coefficients, calculates fluctuation ranges of distortion coefficients, and optimizes visual point clouds and camera poses during the reconstruction process, allowing for accurate 3D reconstruction even without calibrated parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If calibrated intrinsic parameters of the camera are used, then reconstruction accuracy is improved, but the method cannot be applied when camera parameters are not calibrated
Solution Approach 1:
The patent performs preliminary calibration to obtain nominal intrinsic parameters and distortion coefficients before the 3D reconstruction process. These pre-obtained parameters serve as initial values for the optimization algorithm, enabling the system to work with both calibrated and uncalibrated cameras while maintaining reconstruction accuracy.
Solution Approach 2:
The patent treats the camera intrinsic parameters and distortion coefficients as variable parameters that can be optimized during the 3D reconstruction process. By allowing these parameters to fluctuate within certain ranges and be adjusted through optimization, the system adapts to different camera calibration states and improves reconstruction accuracy.
2Adaptability or versatility
If nominal intrinsic parameters and initial distortion coefficients are used without calibration, then adaptability is improved, but reconstruction accuracy deteriorates
Solution Approach 1:
The patent implements an optimization feedback mechanism where the 3D reconstruction results are continuously evaluated and used to adjust the camera parameters and distortion coefficients. This iterative optimization process refines the initial nominal parameters, improving reconstruction accuracy while maintaining the ability to work with uncalibrated cameras.
Solution Approach 2:
The patent transforms static nominal parameters into dynamic variables that can be adjusted during the reconstruction process. By allowing parameters to change and optimize based on the actual image data and geometric constraints, the system achieves both adaptability and accuracy.
3Measurement precision
If optimization is performed during 3D reconstruction, then reconstruction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary preparation by obtaining nominal intrinsic parameters and distortion coefficients before optimization. This pre-processing step provides good initial values that converge faster during optimization, reducing computational complexity while maintaining accuracy improvements.
Solution Approach 2:
The patent applies optimization selectively to specific parameters (camera pose, intrinsic parameters, distortion coefficients) rather than all parameters simultaneously. This localized optimization approach reduces computational complexity while still achieving significant accuracy improvements in the most critical parameters.
Data Source
AI summary
The disclosure provides a 3-dimensional reconstruction method, an electronic device and a storage medium. The method includes: obtaining an image sequence pre-collected by a camera; obtaining nominal intrinsic parameters of the camera, and preset initial values of distortion coefficients of the camera, the camera being a camera for collecting the image sequence; calculating fluctuation ranges of the distortion coefficients; performing the 3D reconstruction on the image sequence based on the nominal intrinsic parameters, the initial values of the distortion coefficients, and the fluctuation ranges of the distortion coefficients; and during the 3D reconstruction, optimizing an obtained visual point cloud and a camera pose.