(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Noise level reduction of 2-d signal using fuzzy logic / Tan Chau Wei

Noise level reduction of 2-d signal using fuzzy logic Tan Chau Wei_E3_2007_NI
Penggunaan sistem pemantauan automatik dalam proses pembuatan telah meningkatkan kepentingan pemprosesan imej digital dalam dunia perindustrian. Sebagai cabang utama pemprosesan imej digital, teknik penurasan hingar yang sempurna semakin ditekankan. Di samping teknik pengurangan hingar tradisional, penyelidik-penyelidik telah menceburi bidang pembangunan teknik pengurangan hingar yang tidak linear. Objektif projek ini ialah untuk merekabentuk penuras digital tidak linear yang berupaya mengurangkan aras hingar dengan berkesan tanpa mengubah ciri-ciri dan tekstur imej asal. Pemprosesan imej fuzzy, merupakan suatu teknik yang mengaplikasikan teori logik fuzzy. Teknik ini mengaplikasikan fungsi keahlian fuzzy ke dalam pemprosesan imej untuk menentukan tindakan yang perlu diambil semasa mengendalikan hingar. Dalam projek ini, suatu teknik pengurangan hingar yang dinamakan “Multipass Fuzzy Filter” telah dicadangkan. Penuras yang dicadangkan merupakan suatu penuras yang dapat mengurangkan hingar denyut dan seragam dengan berkesan di samping mengekalkan ciri-ciri imej asal dengan mengaplikasikan fungsi keahlian hingar denyut dan fungsi keahlian hingar seragam. Penuras ini terdiri daripada tiga bahagian, dengan setiap bahagian berfungsi untuk mengurangkan jenis hingar yang berlainan. Secara umumnya, hingar denyut akan dituras oleh bahagian pertama dan selepas itu, hingar seragam dalam keluaran bahagian pertama akan dituras oleh bahagian kedua. Bahagian ketiga pula akan memproses keluaran daripada bahagian kedua dan bertindak sebagai modul pembetulan. Dalam projek ini, “Mean square error” (MSE) telah digunakan untuk mengukur prestasi penuras-penuras. Keputusan eksperimen telah menunjukkan bahawa penuras fuzzy yang dicadangkan berupaya mengurangkan aras hingar dengan berkesan tanpa mengubah ciri-ciri dan tekstur imej asal. _________________________________________________________________________________________ The increasing popularity of automated visual inspection system for manufacturing process has demonstrated the increasing importance of digital image processing in industrial area. As a major branch in image processing, the needs for perfect digital noise removal filter are receiving more and more attentions. Instead of conventional noise removal techniques, researchers have made their move into non-linear noise removal techniques. The objective of this project is to design a non-linear digital filter that is capable to reduce noise level in digital signals while preserving features and structures of the original signals. Fuzzy image processing, a term that reveals the applications of fuzzy logic into image processing, is an image processing technique in which fuzzy membership functions are used to decide the actions to be taken in handling noise. In this dissertation, a Multipass Fuzzy Filter is proposed. The proposed filter is a feature preserving and perfect noise removal filter, adapting an impulse noise membership function and a proposed uniform noise membership function. The filter consists of three blocks, each block designed with membership functions for different type noise removal. First, the noisy image is processed by the first block where impulse noise is removed. Then, the output is feed to the second block for uniform noise removal. The output is then processed by the third block which acts as a correction module. Mean square error (MSE) has been used to measure the performance of the filter. The experimental results has clearly demonstrated that the proposed fuzzy filter is capable of removing impulse and uniform noise efficiently while preserving details and textures in image that are intensively corrupted.
Contributor(s):
Tan Chau Wei - Author
Language:
English
Subject Keywords:
Digital noise reduction, fuzzy image processing, multipass fuzzy filter
First presented to the public:
3/1/2007
Original Publication Date:
9/6/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 82
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-09-06 16:52:51.062
Date Last Updated
2019-01-07 11:24:32.9118
Submitter:
Nor Hayati Ismail

All Versions

Thumbnail Name Version Created Date
Noise level reduction of 2-d signal using fuzzy logic / Tan Chau Wei1 2018-09-06 16:52:51.062