(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Classification of pet bottles using neural network / Nor Aimi Izzati Mohd Nor

Dengan pemanasan global yang menyebabkan perubahan iklim besar di seluruh pelusuk dunia, telah menjadi tugas utama kita untuk mengambil langkah-langkah tertentu bagi mengurangkan pemanasan global supaya persekitaran kita terpelihara. Kitar semula secara meluas di anggap memberi manfaat kepada alam sekitar, meskipun pengumpulan, pengelasan dan pemprosesan bahan-bahan ini menimbulkan beberapa kesan alam sekitar dan penggunaan tenaga. Oleh kerana plastik adalah bahan yang paling banyak digunakan dalam kehidupan kita sehari-hari, kenyataan masalah di dalam tesis ini tertumpu kepada pengelasan atau klasifikasi dan pemprosesan botol plastik PET dan non-PET. Tesis ini memperkenalkan sebuah sistem pengelasan mesra pengguna (GUI) menggunakan rangkaian saraf buatan berdasarkan ciri-ciri warna dan ciri-ciri bentuk gambar botol. Kami mencadangkan pendekatan untuk mengkelaskan botol PET menggunakan kombinasi pemprosesan imej dan rangkaian saraf buatan. Umumnya,projek ini dapat mengenali botol plastik dikitar semula yang merupakan plastik PET botol. Ciri-ciri yang umum dalam pemprosesan imej melibatkan warna, bentuk dan tekstur. Bagi ciri-ciri warna, keamatan saluran terpisah merah, hijau, biru (RGB) lebih disukai berdasarkan warna dari botol itu sendiri di mana non-PET botol lebih kabur dari PET botol. Hal ini dapat memberikan perbezaan yang jelas dalam keamatan imej kelabu setiap saluran RGB. Bagi ciri-ciri bentuk, sebuah parameter seperti ukurkeliling, luas dan titik tengah dipilih kerana merekan adalah umum dalam pemprosesan imej. Tahap kejayaan pengelasan adalh 89%, membuktikan keberkesanan dan prestasi sistem ini. _________________________________________________________________________________________ With global warming causing major climatic changes in different parts of the world, it becomes our primary duty to take certain steps to reduce it so that the environment can be protected. Recycling is widely assumed to be environmentally beneficial, although the collection, sorting and processing of materials give rise to some environmental impacts and energy use. Since plastic is the most material that used in our life everyday, the problem statement in this study is focusing on classifying and processing of plastic bottles. This thesis introduces a user friendly classification system (GUI) using neural network based on color feature and shape feature of bottle image. We propose an approach to classify PET bottles by combination of image processing and neural network. Generally, this project can recognize a recyclable plastic bottle which is a PET plastic bottle. Common feature of image processing involving color, shape and texture. For color feature, the average intensity of separated red, green, blue (RGB) channels is preferred based on the color of the bottle itself where Non-PET bottle is more opaque than PET bottle. This can give significance different in intensity of grey level for every channel of RGB. For shape feature, a parameter such as perimeter, area and centroid are chosen as they are common in image processing. Its successful classification rate is 89%, which also represents the effectiveness and performance of this system.
Contributor(s):
Nor Aimi Izzati Mohd Nor - Author
Primary Item Type:
Final Year Project
Identifiers:
Accession Number : 875004012
Language:
English
Subject Keywords:
global warming ; PET plastic bottle.; opaque ;
First presented to the public:
4/1/2011
Original Publication Date:
5/31/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 96
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-05-31 15:30:27.618
Date Last Updated
2019-01-07 11:24:32.9118
Submitter:
Nor Hayati Ismail

All Versions

Thumbnail Name Version Created Date
Classification of pet bottles using neural network / Nor Aimi Izzati Mohd Nor1 2018-05-31 15:30:27.618