An object-based approach to detect tree stumps in a selective logging area using Unmanned Aerial Vehicle imagery

Abstract

Acquiring tree-stump information is important to support post-harvest site assessment. Unmanned Aerial Vehicles (UAVs) have been widely used as a tool for analyzing selective logging impacts in forest area sites. One of the potential use of UAV imagery data for analyzing the impact of selective logging is by obtaining tree stump information. Feature extraction and segmentation images to extract stumps from a UAV scene of a forested area in Ulu Jelai, Pahang provides a quick, automated method for identifying stumps. This research implemented a technique for detecting, segmenting, classifying, and measuring tree stumps by using the Multiresolution Segmentation Algorithm method. This study assessed the capability of an object-based approach on image detection to segment and merge the stumps after selective logging practice on UAV imagery with a scale of 0.06-meter resolution. The results revealed that the tree-stumps were detected with an accuracy of 70% and stumps classification were detected with 80% accuracy validated with the ground points. The accuracy is acceptable for data acquiring after 6 months of logging activities. The findings of this study are promising and can lead to increase support for a more cost-effective and systematic selective logging in the future. An effective management system can help related authorities and agencies to develop and maintain the selective logging technique towards sustainable forest management.