Ekstrasi Fitur Bentuk Daun Tanaman Obat Menggunakan Metode Pengolahan Citra

Sari

There are many kind of the medicine plants in Indonesia, that can confuse the people to know them. The leaves are part of the plant that have different pattern and shape from one leaf to another leaf. In this research is using canny edge detection operator and backpropagation artificial neural network so the medicine plants can be recognized. Edge detection is one of the methods of image processing that capable to identify an object, because edge detection can characterize the boundaries of an object. Artifical neural network is an example of the pattern recognition that can work like human brain to get knowledges and some informations. Based on the results of some experiments that have been done, the results showed that using trained data with 95,71% accuracy while experiment that using untrained data with 80% accuracy, and also implementation in realtime with 73,3% accuracy.