Hyperaccurate Correction of Maximum Likelihood for Geometric Estimation
スポンサーリンク
概要
- 論文の詳細を見る
The best known method for optimally computing parameters from noisy data based on geometric constraints is maximum likelihood (ML). This paper reinvestigates "hyperaccurate correction" for further improving the accuracy of ML. In the past, only the case of a single scalar constraint was studied. In this paper, we extend it to multiple constraints given in the form of vector equations. By detailed error analysis, we illuminate the existence of a term that has been ignored in the past. Doing simulation experiments of ellipse fitting, fundamental matrix, and homography computation, we show that the new term does not effectively affect the final solution. However, we show that our hyperaccurate correction is even superior to hyper-renormalization, the latest method regarded as the best fitting method, but that the iterations of ML computation do not necessarily converge in the presence of large noise.
著者
-
Sugaya Yasuyuki
Department Of Computer Science And Engineering Toyohashi University Of Technology
-
Kanatani Kenichi
Department Of Computer Science Gunma University
関連論文
- Geometric BIC
- Uncalibrated Factorization Using a Variable Symmetric Affine Camera(Image Recognition, Computer Vision)
- Statistical Optimization for 3-D Reconstruction from a Single View(Image Recognition and Understanding)
- Do We Really Have to Consider Covariance Matrices for Image Feature Points?
- Optimal Conic Fitting and Reliability Evaluation
- Infinity and Planarity Test for Stereo Vision
- Direct Reconstruction of Planar Surfaces by Stereo Vision
- Reliability of Fitting a Plane to Range Data
- Optimal Line Fitting and Reliability Evaluation
- Optimizing a Triangular Mesh for Shape Reconstruction from Images(Image Recognition and Understanding)
- Outlier Removal for Motion Tracking by Subspace Separation
- Extending Interrupted Feature Point Tracking for 3-D Affine Reconstruction(Image Recognition, Computer Vision)
- Optimization without Minimization Search : Constraint Satisfaction by Orthogonal Projection with Applications to Multiview Triangulation
- Multi-Stage Unsupervised Learning for Multi-Body Motion Segmentation(Image Recognition, Computer Vision)
- Reliability of 3-D Reconstruction by Stereo Vision
- High Accuracy Fundamental Matrix Computation and Its Performance Evaluation(Image Recognition, Computer Vision)
- Ellipse Fitting with Hyperaccuracy(Image Recognition, Computer Vision)
- Renormalization for Motion Analysis: Statistically Optimal Algorithm (Special Issue on Computer Vision)
- Optimal Structure-from-Motion Algorithm for Optical Flow
- Optimal grid pattern for focal length calibration
- Hyperaccurate Correction of Maximum Likelihood for Geometric Estimation
- Hyperaccurate Correction of Maximum Likelihood for Geometric Estimation
- Optimal Robot Self-Localization and Accuracy Bounds