Identifikasi Kualitas Kayu Berdasarkan Pada Analisa Suara Dan Neural Network

Wachid, Abdur Rochman (2016) Identifikasi Kualitas Kayu Berdasarkan Pada Analisa Suara Dan Neural Network. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
2212204015-Master_Thesis.pdf - Accepted Version

Download (1MB) | Preview

Abstract

industri kayu terdapat beberapa jenis kayu dengan kualitas yang beragam. Kualitas kayu dapat dibedakan dari tekstur, kelembaban serta kepadatannya. Kualitas kayu tidak dapat diukur melalui penglihatan atau sentuhan, karena kamera hanya dapat mendeteksi permukaan luar kayu. Untuk itu dibutuhkan suatu cara atau metode yang tepat untuk membedakannya. Pada penelitian ini telah dibuat suatu pengukuran kualitas kayu berbasis suara. Kualitas kayu dibedakan berdasarkan suara dari pukulan pada kayu. Suara dirubah dari ranah waktu menjadi frekuensi menggunakan algoritma Fast Fourier Transform. Deret komponen frekuensi tersebut digunakan sebagai masukkan pada algoritma Neural Network yang dilatih untuk mengetahui jenis dan kualitas dari kayu. Pada percobaan ini digunakan tiga jenis kayu, yaitu kayu jati, meranti, dan kruing. Sedangkan pada pengujian kualitas kayu digunakan dua jenis kayu dengan masing-masing mempunyai dua kualitas yang berbeda. Hasil percobaan menunjukkan bahwa Neural Network dapat mengenali setiap jenis kayu dengan taraf keberhasilan 93,33%. Sedangkan untuk kualitas kayu, sistim ini dapat mengenali setiap kualitas kayu dengan taraf keberhasilan 95%. ===================================================================================================== In the timber industry, there are several types of wood with varying quality. The quality of wood can be distinguished from the texture, moisture and density. Wood quality can not be measured through sight or touch, because the camera can only detect the outer surface of the wood. Therefore it requires a proper way or method to distinguish. This study has been made of a sound based quality measurement. Wood quality was differentiated by the sound of blows on the wood. Sound was converted from the time domain into the frequency using Fast Fourier Transform algorithm. The series of frequency components were used as an input of the Neural Network algorithm trained to discriminate the type and quality of woods. In this experiment, there were three types of woods, the teak, meranti, and kruing. While in testing the quality of the wood was used two types of woods with each having two different qualities. The experimental results showed that the Neural Network could identify any type of woods with the level of success of 93.33%. As for the quality of wood, this system could identify any quality of woods with a level of success of 95%

Item Type: Thesis (Masters)
Additional Information: RTE 006.3 Wac i
Uncontrolled Keywords: Fast fourier transform, Kualitas kayu, Sinyal suara, Neural network
Subjects: Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Divisions: Faculty of Industrial Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User: Yeni Anita Gonti
Date Deposited: 28 Apr 2020 06:51
Last Modified: 28 Apr 2020 06:51
URI: https://repository.its.ac.id/id/eprint/75900

Actions (login required)

View Item View Item