Ekstraksi Fitur Bentuk dan Venasi Citra Daun dengan Pemodelan Fourier dan B-Spline

Abstract

Leaf features extraction on plant leaf image is still be a problem on automatic plant leaf identification. Shape and venation of leaf are a significant part of leaf for distinguishing a species of leaf from each other. Shape feature extraction with Hough transform and Fourier descriptor were implemented and their effectiveness on leaf shape recognition were compared. Effectiveness of both shape recognition methods were evaluated by recall-precision measurement. Recall-precision evaluation showed that leaf shape represented by Fourier descriptors is more effective than represented by Hough transform. Plant leaf image vein extraction using b-spline representation was implemented. An automatic initialitation of vein search parameter using Standard I-Iough Transform was proposed.