.

ISSN 2063-5346
For urgent queries please contact : +918130348310

CaragaITreePier: A Mobile Application for Wood Identification Using Deep Learning

Main Article Content

Shekinah Mae E. Zapa , Avemelic D. Cabanalan , Judy Ann S. Guitguitin , Rey N. Cossid , Roselyn L. Palaso , Roger T. Sarmiento
» doi: 10.48047/ecb/2023.12.si4.387

Abstract

The wood industry is vibrant in Mindanao, and particularly in the Caraga Region. Industrial tree plantation (ITP) has become a solution to declining forest cover, as the extraction of wood from natural forests was gradually regulated until eventually banned in 2010. Identification of the wood of ITP species is a valuable process. From a traditional method of identifying wood and preventing fraud during timber trading, the authors envisioned a modern approach that is accurate and hassle-free by integrating knowledge in forestry and the advancement of technology. Therefore, CaragaITreePier-a mobile-based application that uses deep learning algorithm-was developed for the wood identification of priority ITP species in the region. With this application, ordinary people can now perform proper wood identification by simply scanning the wood using macrolens attached to a cellphone. The results of the application evaluation showed a high precision in the classification of wood samples. In terms of performance efficiency, the mobile application provided users a smooth experience in loading, predicting, and displaying information. Finally, the overall evaluation of the user respondents further confirmed that CaragaITreePier is user-friendly and a promising tool for the classification and authentication of wood species. The fundamental concept of the system is a trained artificial intelligence model with deep learning capabilities that can match human-level accuracy while using computing power for wood identification. Moreover, the advantage of the system is its portability, as CaragaITreePier can be used on mobile phones anywhere even without internet connectivity.

Article Details