Point Quality Evaluation Metric (PQM) for LiDAR-Based Mapping and Reconstruction

A novel metric to comprehensively assess the quality of large point-clouds from LiDAR-based mapping.

Background:

LiDAR-based mapping and reconstruction are crucial for various applications, yet assessing the quality of dense maps they generate poses challenges.  Current methods often fail to capture completeness, structural information, and local variations in error.

Technology Overview:

The Point Quality Evaluation Metric (PQM) developed by University at Buffalo researchers introduces a novel approach to evaluate the quality of large point-clouds.  PQM consists of four sub-metrics:

• Completeness evaluates the proportion of missing data 
• Artifact Score recognizes and characterizes artifacts
• Accuracy measures registration accuracy
• Resolution quantifies point cloud density

https://buffalo.technologypublisher.com/files/sites/7566_in-part.jpg

Source: Osteira, https://stock.adobe.com/uk/825705197, stock.adobe.com

Advantages:

Comprehensive, normalized, efficient, GPU-accelerated, scalable, and structure-aware assessment of point cloud quality

Applications:

• Sensor evaluation tools
• Anomaly/change detection
• Quality estimation
• Augmented reality/virtual reality
• Autonomous driving
• Mobile robotics
• Autonomous inspection/infrastructure analysis

Intellectual Property Summary:

Provisional patent application 63/657,108 filed June 6, 2024.

Stage of Development:

TRL 6

Licensing Status:

Available for licensing or collaboration.

Figures:

https://buffalo.technologypublisher.com/files/sites/7566_image_2024-07-15_113839.jpg