Back to Search
Start Over
Wood Species Image Classification Using Two-Dimensional Convolutional Neural Network
- Source :
- Drvna industrija; ISSN 0012-6772 (Print); ISSN 1847-1153 (Online); Volume 74; Issue 4
- Publication Year :
- 2023
-
Abstract
- The woodworking industry’s recognition and classification of timber is essential for trade, production and timber science. Traditional methods of identifying wood types are complex, time-consuming, costly and require expertise in wood science. Traditional techniques have been replaced by convolutional neural networks (CNNs), a deep learning tool to better identify wood species. In contrast to earlier studies that used pretrained models, a novel architecture designed explicitly for the WOOD-AUTH dataset was proposed in this study to develop a new 2D CNN model. The data collection encompasses high-level visual representations of 12 distinct types of timber. It is aimed to create a simpler and faster model as an alternative to time-consuming and heavy wood classification models. Compared to previous studies, this research worked with a newly structured 2D CNN network based on 12 wood species. High accuracy and fast computation time were achieved using fewer numbers (three layers) of the convolutional neural network. The proposed model achieved 94 % accuracy, 87 % precision, 81 % recall, 80 % F1 score and 112 minutes 27 seconds computation time. The 2D CNN model performed better than the transfer learning models regarding training epochs. The primary benefit of the model is its ability to achieve high accuracy with lower computation time, even at high epochs compared to other models. The introduced 2D CNN model produced satisfactory outcomes for wood species classification.<br />Identifikacija i klasifikacija drva u drvnoj industriji ključna je za trgovinu, proizvodnju i znanost o drvu. Tradicionalne metode identifikacije vrste drva složene su, dugotrajne i skupe te zahtijevaju stručnost s područja znanosti o drvu. Za bolju identifikaciju vrste drva tradicionalne su metode zamijenjene konvolucijskim neuronskim mrežama (CNN), odnosno alatom za duboko učenje. Za razliku od ranijih studija koje su se koristile unaprijed obučenim modelima, u ovoj je studiji predložena nova arhitektura dizajnirana upravo za skup podataka WOOD-AUTH kako bi se razvio novi 2D CNN model. Zbirka podataka obuhvaća vizualne prikaze visoke razlučivosti 12 različitih vrsta drva. Cilj je bio stvoriti jednostavniji i brži model kao alternativu dugotrajnim i složenim modelima klasifikacije drva. Za razliku od prethodnih istraživanja, u ovom je istraživanju primijenjena nova 2D CNN mreža koja se temelji na 12 vrsta drva. Visoka točnost i brzo vrijeme izračuna postignuti su korištenjem manjeg broja slojeva (tri sloja) konvolucijske neuronske mreže. Predloženim je modelom postignuta točnost od 94 %, preciznost od 87 %, opoziv od 81 %, F1 rezultat od 80 % i vrijeme izračuna od 112 minuta i 27 sekundi. Model 2D CNN pokazao se boljim od modela transfernog učenja u smislu epohe poduke. Primarna prednost modela jest njegova sposobnost postizanja visoke točnosti uz kraće vrijeme izračuna, čak i pri visokim epohama u usporedbi s drugim modelima. Prezentirani 2D CNN model dao je zadovoljavajuće rezultate za klasifikaciju vrste drva.
Details
- Database :
- OAIster
- Journal :
- Drvna industrija; ISSN 0012-6772 (Print); ISSN 1847-1153 (Online); Volume 74; Issue 4
- Notes :
- application/pdf, English
- Publication Type :
- Electronic Resource
- Accession number :
- edsoai.on1422853977
- Document Type :
- Electronic Resource