Efficient and Deterministic Search Strategy Based on Residual Projections for Point Cloud Registration With Correspondences

Keyphrases

• 2D-3D Correspondence 33%
• 3D Feature Matching 33%
• Algorithm Efficiency 33%
• Branch-and-bound 66%
• Branch-and-bound Optimization 33%
• Challenging Problems 33%
• Comprehensive Experiment 33%
• Computational Complexity 33%
• Decoupling Strategy 33%
• Deterministic Search 100%
• Dimensional Parameters 33%
• Efficient Search 100%
• Interval Stabbing 33%
• LiDAR Scanning 33%
• Matching Approach 33%
• Optimization Framework 33%
• Original Problem 33%
• Outlier-robust 33%
• Parameter Space 33%
• Point Cloud Registration 100%
• Registration Method 33%
• Registration Problem 66%
• Residual Projection 100%
• Rigid Transformation 33%
• Robust Registration 33%
• Search Space 33%
• Search Strategy 100%
• Six Degrees of Freedom 33%
• Solution Domain 33%
• State-of-the-art Techniques 33%
• Three Degrees of Freedom 33%
• Two-dimensional Domain 33%
• Typical Points 33%

Engineering

• Computational Complexity 33%
• Conducted Experiment 33%
• Degree of Freedom 66%
• Dimensional Domain 33%
• Parameter Space 33%
• Point Cloud 100%
• Rigid Transformation 33%
• Search Space 33%
• Search Strategies 100%
• Solution Domain 33%
• State-of-the-Art Method 33%
• Subproblem 100%
• Two Dimensional 33%

Computer Science

• Branch-and-Bound Algorithm Design 100%
• Computational Complexity 33%
• Dimensional Parameter Space 33%
• Matching Approach 33%
• Optimization Framework 33%
• Point Cloud 100%
• Rigid Transformation 33%
• Search Space 33%
• Search Strategies 100%
• Solution Domain 33%