Priyono,, Aziz Hari (2017) MARINE WASTE CLASSIFICATION USING MOMENT INVARIANTS AND NAÏVE BAYES CLASSIFIER. S1 thesis, Universitas Mataram.
Text
PAPER.docx Download (3MB) |
Abstract
People have transformed marine areas into huge garbage container where other creatures are feeding on it. Plastics as one of major waste produced by human have contaminated food chain. To date, there are no cheap or simple method for managing the waste. In this work, Naïve Bayes Classifier combined with Moment Invariants system is developed to help classifying floating waste on marine areas. This system is implemented in Java by using 2000 sample data from marine and several experimental environments. Several image processings are also used such as resizing, Otsu Thresholding and Histogram Equalization. The results obtained from the proposed system are acceptable in accuracy (69.35%) and False Positive Rate (20.52%) but unreliable False Negative Rate (61.06%). This results are due to overlapping features distribution produced from 10 ranges of Moment Invariants. Although the results are still far from good, the proposed method opens limitless improvements for the next implementations.
Item Type: | Thesis (S1) |
---|---|
Keywords (Kata Kunci): | Naïve Bayes Classifier, Moment Invariant, Image Processing, Image Classification, Marine Waste |
Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering |
Divisions: | Fakultas Teknik |
Depositing User: | Ayus Suyarsih |
Date Deposited: | 31 Jul 2018 06:25 |
Last Modified: | 31 Jul 2018 06:25 |
URI: | http://eprints.unram.ac.id/id/eprint/7043 |
Actions (login required)
View Item |