Classification of Wood Type on Android based Building Construction

ABSTRACT There are many types of wood in Indonesia, so it is difficult to remember everything. Various types of wood can be used as building materials and household furniture, to find out the type of wood that can be used as a specific material and to help people in determining its type, then we need a system that can be used to classify wood types quickly and accurately . The applications can run on Android device and it is created using Local Binary Pattern (LBP) method as a feature extraction process and Support Vector Machine (SVM) with linear kernel implementation as a classification method. This application runs using web services with a separate database located on the server which is called RESTful web service, so it doesn't take up much storage space on an Android device . The system testing that has been carried out gives the classification results with the highest level of accuracy is 84%, macro precision is 84.55%, macro recall is 84% and macro-F1 is 83.65%. Keywords: Classification, Wood Type, Local Binary Pattern (LBP), Support Vector Machine (SVM), Android.