Wood Quality Classification Based on Texture and Fiber Pattern Recognition using HOG Feature and SVM Classifier

Wood as a material for household appliance needs to be considered of quality. Quality of wood can be classified according to colours, texture, and wood fibre pattern differences. In general, wood industries have been doing the wood quality classified process using a conventional method with a sense of vision in which the results are subjective in terms of accuracy and time efficiency. Machine Learning is a solution to this problem of predicting and classifying data of wood quality. In this paper, the wood will be recognized using Histogram of Oriented Gradient to know the pattern and texture. Meanwhile, the classification method uses Support Vector Machine which will be compared to find the best accuracy and time computation. This system is given image input with five types of cedar classification such as Class A, Class B, Class C, Class D, and Class E took using Logitech C930e HD which is integrated with Arduino Uno for object detection process and conveyor. The Experiment achieve 90% of accuracy with time computation 1,40 s