Analisis Filtering Noise Menggunakan Metode Statistik Mean, Median pada Objek Citra Malam Hari
No Thumbnail Available
Date
2017-06-02
Authors
Rahmad, Rahmad
Journal Title
Journal ISSN
Volume Title
Publisher
Politeknik Negeri Batam
Abstract
Image filtering is one part of image quality improvement, by smoothing and
eliminating noise on the image, both linearly and non-linearly. The mean filter is
one linear filtering that works by replacing the pixel value intensity with the
average of the pixel value with its neighboring pixel values. While the median
filter is one of the non-linear filtering that sort the intensity value of a pixel group,
then replace the pixel value that is processed with its median value. This study
studied noise reduction techniques using the mean and median filter method, to
reduce the noise subsequently conducted in a trial of multiple color images with
30 previously selected image seals, Test image is given the noise of Salt & Pepper
shaped like black spots and White on the image to reduce the noise is required
method of noise filtering using median and mean filter. Then analyzed the
performance qualitatively by comparing the image output filter, image bernoise
and original image by naked eye. Quantitatively measured through PSNR, SNR
and MSE values. Testing more in depth on the performance of this filter by
finding the value of the difference between each value of PSNR, SNR and MSE
and get good results. The difference is in the value of PSNR, SNR and MSE and
the difference that has been obtained through filtering noise. It is concluded that
good noise filtering using median filter method mean method is not good to be
used in testing noise filtering on image.
Description
Keywords
Teknik Informatika, Multimedia & Jaringan, Ilmu Komputer, Pengetahuan, Sistem, Analysis, Filtering, Statistical Methods