A Comparative Study of Feature Extraction Methods for Wood Texture Classification

The objective of this paper is to evaluate the classification performance of several feature extraction and classification methods for exotic wood texture images as dataset. The Gray Level Co-occurrence Matrix, Local Binary Patterns, Wavelet, Ranklet, Granulometry, and Laws' Masks will be used to extract features from the images. The extracted features are then fed into five classification techniques: Linear and Quadratic Classifier, Neural Networks, Support Vector Machine, and K-Nearest Neighbor so that each class membership can be obtained. The success rate of each method then measured by comparing the predicted labels and its ground truth so that in the end, the best feature extraction method will be indicated by the highest classification rate. This paper provides recommendations in feature extraction method and classification technique which may give good result in similar task. By considering several factors, such as: computational complexity, classification rate, and running time, this work has found that LBP is more appropriate to analyze wood texture.