Pine and spruce roundwood species classification using multivariate image analysis on bark

Abstract

Wood discs from 67 pine and 79 spruce logs were collected from a forest clearing. Three different 24-bit redgreen-blue (RGB) images were acquired from the radial surface of each disc. The first image contained bark, the second image was a mixture of bark and wood surface, and the third image consisted only of wood surface. The image texture was compressed into vectors of Fouriertransformed wavelet coefficients. These were assembled in matrices and analysed by principal component analysis (PCA) and partial least-squares projections to latent structures (PLS). Classification using Fourier-transformed wavelet scales showed that the wood species could be predicted with 90% accuracy. A thorough examination of this classification showed that the predicting power of these models was mostly due to wavelet scales that represented the mean value of each colour channel. The prediction accuracy that could be obtained from coefficients representing image texture was generally low. The use of grey-level co-occurrence matrices prior to the wavelet transformation showed, however, that it is possible to classify the wood species of pine and spruce with an accuracy approaching 100%.

Image analysis of weaverbird nests reveals signature weave textures

“…We chose the five methods based on their success in other contexts (classification of coral-reef images: [ 32 ], fabric defect detection: [ 33 ], e.g. tree bark species identification: [ 34 ], plant leaf identification: [ 35 ]) to maximize the likelihood that one of these methods would prove useful for classifying nest weave texture. For each dataset, we then tested the value of a variety of different combinations of textural descriptors for classifying nest images.…”

A randomization test for PLS component selection

“…Since most of these data sets are detailed elsewhere [56][57][58][59][60][61][62][63][64],only a short description follows. Table I gives an overview and also indicates how four data sets are further divided in a calibration and validation set in order to present a final test of the different component selection rules.…”

Detecting spruce and fir log diameter under bark by using shape and image data

“…3D scanners), so the tracheid effect can easily be used for bark detection if some cameras for filming the laser line are added. Colour images of logs can also be used to distinguish bark and wood if suitable image processing algorithms are applied (Nilsson & Edlund 2005;Denzler et al 2013). For documentation purposes, imaging systems are already in use in many sawmills.…”