Font Size: 

To improve the reliability of furniture, residential, and other wood products, wood traceability technology is needed [1]. Traditionally, the wood can be tracked by attaching a barcode or IC tag. But there are several risks to traditional traceability technology such that the IC tag is pulled out illegally or the barcode is damaged during drying or processing.In this paper, we propose to use a tree-ring image to recognize the timber because the tree-ring has a special pattern and cannot be forged by hand. Learning models are performed by using a deep learning algorithm of convolution neural network (CNN) [2]. The tree-ring image labeled with a number is used as an input vector. We confirmed the effectiveness of timber recognition with a tree-ring image using CNN.

Convolution neural network (CNN), Timber traceability, Timber industry