Deep Learning methodology for the identification of wood species using high-resolution macroscopic images

TL;DR

The study tackles illegal timber trade and regulatory compliance by addressing timber species identification from high-resolution macroscopic images. It introduces a patch-based TDLI-PIV framework that trains CNNs on patches and uses a patch-wise voting scheme during inference, achieving state-of-the-art accuracy on a new high-resolution GOIMAI-Phase-I dataset (2120 images, 37 species). Key findings show that patching preserves fine-grained timber patterns better than full-image downscaling, with TDLI-PIV reaching 99.4% accuracy and robust performance even with only 25% of the data. The work also provides a public high-resolution dataset and demonstrates practical viability for border-control contexts, while outlining challenges in standardizing data collection across devices and magnifications.

Abstract

Significant advancements in the field of wood species identification are needed worldwide to support sustainable timber trade. In this work we contribute to automate the identification of wood species via high-resolution macroscopic images of timber. The main challenge of this problem is that fine-grained patterns in timber are crucial in order to accurately identify wood species, and these patterns are not properly learned by traditional convolutional neural networks (CNNs) trained on low/medium resolution images. We propose a Timber Deep Learning Identification with Patch-based Inference Voting methodology, abbreviated TDLI-PIV methodology. Our proposal exploits the concept of patching and the availability of high-resolution macroscopic images of timber in order to overcome the inherent challenges that CNNs face in timber identification. The TDLI-PIV methodology is able to capture fine-grained patterns in timber and, moreover, boosts robustness and prediction accuracy via a collaborative voting inference process. In this work we also introduce a new data set of marcroscopic images of timber, called GOIMAI-Phase-I, which has been obtained using optical magnification in order to capture fine-grained details, which contrasts to the other datasets that are publicly available. More concretely, images in GOIMAI-Phase-I are taken with a smartphone with a 24x magnifying lens attached to the camera. Our data set contains 2120 images of timber and covers 37 legally protected wood species. Our experiments have assessed the performance of the TDLI-PIV methodology, involving the comparison with other methodologies available in the literature, exploration of data augmentation methods and the effect that the dataset size has on the accuracy of TDLI-PIV.

Figures (6)

• Figure 1: Set-up to capture macroscopic images of timber.
• Figure 2: Examples of high-resolution images of timber from different wood species in the GOIMAI-Phase-I dataset.
• Figure 3: Number of images in each class of the database.
• Figure 4: The TDLI-CPIV methodology.
• Figure 5: Comparison between the original image (a) and the cropped image (b).