Automatic stem mapping by merging several terrestrial laser scans at the feature and decision levels

Area of Science:

• Forestry
• Remote Sensing
• Geomatics

Background:

• Accurate forest inventory relies on detailed ground reference data, traditionally gathered manually, which is labor-intensive and limits attribute collection.
• Terrestrial Laser Scanning (TLS) offers efficient plot-level stem mapping, with single-scan methods facing occlusion issues and multi-scan methods requiring manual registration.
• Current TLS techniques present challenges in efficiency and automation for comprehensive forest plot mapping.

Purpose of the Study:

• To introduce and evaluate a novel Multi-Single-Scan (MSS) method for automated forest plot mapping.
• To overcome the limitations of single-scan occlusion and multi-scan registration complexities in TLS-based forest inventories.
• To achieve accurate and efficient stem detection and mapping in dense forest environments.

Main Methods:

• Developed a Multi-Single-Scan (MSS) approach that processes individual scans automatically.
• Implemented automated detection and merging of stem positions from multiple independent scans.
• Avoided the need for artificial reference targets and point-level registration in the data processing pipeline.

Main Results:

• The MSS method significantly enhanced stem-detection accuracy compared to the traditional single-scan approach.
• Mapping accuracy achieved with the MSS method was comparable to the conventional multi-scan technique.
• The proposed method demonstrated successful application across five dense forest plots without manual intervention for registration.

Conclusions:

• The Multi-Single-Scan (MSS) method provides a robust and automated solution for accurate forest plot mapping using TLS.
• This technique offers a practical alternative to existing methods, improving efficiency and reducing labor in forest inventories.
• The MSS method effectively addresses occlusion and registration challenges, paving the way for more accessible and precise forest data collection.