Analisis kualitas reduksi derau pada citra menggunakan gaussian filter, bilateral filter, dan wiener filter (studi kasus: citra kurang cahaya/malam hari)

Tanata, Elbert (2023) Analisis kualitas reduksi derau pada citra menggunakan gaussian filter, bilateral filter, dan wiener filter (studi kasus: citra kurang cahaya/malam hari). Bachelor thesis, Universitas Pelita Harapan.

[img]
Preview
Text (Title)
Title.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (65kB) | Preview
[img]
Preview
Text (Abstract)
Abstract.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (251kB) | Preview
[img]
Preview
Text (ToC)
ToC.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (562kB) | Preview
[img]
Preview
Text (Chapter1)
Chapter 1.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (456kB) | Preview
[img] Text (Chapter2)
Chapter 2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[img] Text (Chapter3)
Chapter 3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (739kB)
[img] Text (Chapter4)
Chapter 4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (4MB)
[img] Text (Chapter5)
Chapter 5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (185kB)
[img]
Preview
Text (Bibliography)
Bibliography.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (277kB) | Preview
[img] Text (Appendices)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Share Alike.

Download (1MB)

Abstract

Dalam penelitian ini telah dilakukan analisis atas perbandingan 3 jenis filter dalam meningkatkan kualitas citra hasil pemotretan. Analisis ini dilakukan untuk meningkatkan kualitas citra yang dipotret pada malam hari atau pada keadaan minim cahaya akan menimbulkan derau (noise) pada citra yang dapat menurunkan kualitas citra itu sendiri. Derau dapat disebabkan ketika proses konversi citra analog menjadi digital, memperoleh atau menyimpan citra digital atau saat mengirim citra. Filter Gaussian, filter Bilateral, dan filter Wiener dipilih untuk mengurangi derau dan dianalisis performa dari masing-masing filter. Eksperimen dilakukan dalam 2 tahap, yang pertama untuk melihat performa masing-masing filter terhadap data buatan yang ditambahkan derau simulasi, yaitu derau Gaussian, Poisson, Aditif, dan Speckle, untuk mengetahui karakteristik dari masing-masing filter. Eksperimen dilakukan dengan nilai parameter yang berbeda, filter Gaussian dan Bilateral menggunakan parameter standar deviasi atau sigma dan filter Wiener menggunakan parameter besaran piksel tetangga. Eksperimen dilanjutkan dengan menghitung dan membandingkan nilai Peak Signal-to-Noise Ratio (PSNR) dan Signal-to-Noise Ratio (SNR) untuk menghitung rasio dari sinyal pada citra terhadap derau yang memiliki satuan desibel (dB), dan Mean Squared Error (MSE) untuk menghitung kesalahan kuadrat rata-rata pada citra. Setelah mengetahui karakteristik dari masing-masing filter, dilakukan eksperimen kedua menggunakan citra dari lapangan dunia nyata yang dipotret dan dikumpulkan secara manual pada malam hari atau keadaan minim cahaya menggunakan telepon pintar. Lalu, filter digunakan untuk mengurangi derau pada citra dan melakukan perhitungan SNR sebelum dan sesudah filter digunakan. Hasil dari eksperimen pertama menunjukkan filter Bilateral dan Gaussian merupakan filter yang memiliki performa lebih bagus dalam mengurangi derau. Namun pada eksperimen kedua menggunakan data dunia nyata menunjukkan filter Wiener memiliki performa lebih baik. Terdapat 7 data sampel dari pemotretan lapangan dunia nyata. Dari hasil analisis yang dilakukan, filter Wiener dengan kernel piksel tetangga sebesar 6*6 menghasilkan hasil yang paling memuaskan dengan menaikkan nilai SNR 5 dari 7 sampel citra yang dianalisis. Sementara filter Gaussian dan filter Bilateral dimana keduanya menaikkan 4 dari 7 sampel citra./In this research, an analysis has been carried out on a comparison of 3 types of filters in improving the quality of images taken at night or in a low light condition, which can produce noise in the image that reduces the image quality. Noise can be caused when converting images from analog to digital, or when obtaining, storing, or sending images. Gaussian filter, Bilateral filter, and Wiener filter were chosen to be analyze the performance of each filter in reducing noise. The experiment is carried out in 2 stages, the first was to observe at the performance of each filter on dummy data added by simulated noise, namely Gaussian noise, Speckle noise, Poisson noise, and Additive noise. Experiments were carried out with different parameter values, Gaussian and Bilateral filter uses standard deviation or sigma parameters and Wiener filter using neighboring pixel size parameters. Then, the experiment is continued by calculating and comparing PSNR and SNR values to calculate the signal-to-noise ratio of the image which has units of decibel (dB), and MSE to calculate the mean squared error of the image. After the characteristics of each filter are known, the second experiment using real-world data that is taken and collected manually on night or low light condition using smartphone. Then, filters are applied to reduce noisy images and make SNR calculation before and after filters applied. The result of the first experiment concludes that both Bilateral filter and Gaussian filter have better performance on reducing noise. While on the second experiment that uses real-world data shows that Wiener filter had better performance on reducing noise. There are 7 real-world images sample data. The analysis concludes that Wiener filter with 6*6 neighboring pixel kernel has the most satisfying results, by increasing the SNR values 5 out of 7 image samples analyzed. While both Gaussian filter and Bilateral filter increase the SNR values 4 out of 7 image samples.

Item Type: Thesis (Bachelor)
Creators:
CreatorsNIMEmail
Tanata, Elbert01082190019tanatabert18@gmail.com
Contributors:
ContributionContributorsNIDN/NIDKEmail
Thesis advisorTjahyadi, Hendra0410076901hendra.tjahyadi@uph.edu
Uncontrolled Keywords: image processing, image enhancement, noise, SNR, night image
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics
Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics
Depositing User: Elbert Tanata
Date Deposited: 28 Jun 2023 08:50
Last Modified: 28 Jun 2023 08:50
URI: http://repository.uph.edu/id/eprint/56256

Actions (login required)

View Item View Item