Matched Pattern System of Jatropha curcas Fruits by Using Back Propagation (BP)  

Oleh:

Effendi, Zulham ; Ramli, Rizauddin ; Ghani, Jaharah Abdul

Jenis: Article from Proceeding
Dalam koleksi: APCOMS 2009: The 2nd Asia-Pacific Conference on Manufacturing System: Reconfigurable Manufacturing System for Facing Turbulent Manufacturing Environment, November 4th-5th, 2009, Yogyakarta, Indonesia, page IX.27-33.
Topik: Jatropha Curcas Fruits; Matched Pattern; Back Propagation; Neural Networks; Maturity; Recognition
Fulltext: APCOMS GIX-5.pdf (332.4KB)

Isi artikel

This paper presents a matched pattern system of Jatropha curcas fruits by using back propagation (BP). Jatropha curcas is widely accepted as a favorite agricultural solution in subtropical and tropical continents that can produce high quantity and quality feedstock for bio energy. Jatropha curcas oil is categorized as non edible; therefore its availability will not be competed with production of food, consequently become a bright prospect for biodiesel feedstock. Traditional identification of Jatropha curcas fruits are performed by human experts. The quality of the fruit depends on type of defects, skin color and size of the fruit. This research develops a pattern recognition system to identify the Jatropha curcas fruit maturity into various quality categories. The system is divided into a training stage to extract the fruit characteristics and an analyzing stage to recognize the pattern by using the characteristics derived from the training stage. Back propagation diagnosis model is used to recognize the Jatropha curcas fruits pattern. The model is able to recognize the maturity of Jatropha curcas fruits by three levels of grading (raw, ripe, over ripe) with twenty-seven images of Jatropha curcas fruits. At the end of the training, it can be ascertained that our proposed back propagation can achieved its performance function.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali