Identifikasi motif berbasis citra menggunakan wavelet dengan klasifikasi probabilistic neural network (Studi kasus: Direktorat Jenderal Hak Kekayaan Intelektual)

Abstract

Motif is one creation that should be respected and protected. The protection must be on target so nobody is aggrieved. The Directorate General of Intellectual Property Rights (DGIP) is a government institution in charge of implementing that protection. A process or step taken is examining the registered creation. The examination is still done manually, so DGIP spends a lot of time. This research aims to create a design and implementation of a system to identify if the motifs have been registered at DGIP with respectively digital image as the input. Detection process begins with the extraction process using the Haar wavelet to get the characteristics of the image and classification process by using the Probabilistic Neural Network (PNN). This study uses 450 training images divided into 50 classes. Testing uses 200 images consisting of 4 images for each class. The accuracies of the motif identification system are 65% and 64% for Haar wavelet level 2 and level 3, respectively.