DWC (Deep Wood Classifier): A Novel Wood Species Classification Framework Based on Deep Learning
DREWNO, vol.1, no.1, pp.1-19, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Volume: 1 Issue: 1
- Publication Date: 2026
- Doi Number: 10.53502/wood-210419
- Journal Name: DREWNO
- Journal Indexes: Scopus, Science Citation Index Expanded (SCI-EXPANDED), Directory of Open Access Journals
- Page Numbers: pp.1-19
- Open Archive Collection: AVESIS Open Access Collection
- Yozgat Bozok University Affiliated: Yes
Abstract
DWC (Deep Wood Classifier) is a hybrid method that aims to achieve high accuracy and efficiency in wood species classification by combining deep learning and classical machine learning algorithms. In this method, convolutional neural network (CNN) models such as EfficientNetV2B3, Xception, and InceptionResNetV2 are optimised and trained to classify wood species. The accuracy rate is further improved when the features extracted from these deep learning models are classified with classical machine learning algorithms. The combination of EfficientNetV2B3 and SVC provides fast and effective classification with 99.56% accuracy, while Xception and Logistic Regression achieved the highest success with 99.69% accuracy. The DWC method exhibited excellent results in confusion matrix and ROC curve analyses, providing higher accuracy and more efficient training processes compared to existing methods in the literature. The combination of deep learning and classical machine learning algorithms has made DWC stand out with its high accuracy rates and fast training times. This hybrid approach offers a significant innovation in wood species classification, demonstrating superior performance compared to other methods in the field.