Perancangan Aplikasi Perbaikan Citra Digital Pada Hasil Screenshot Dengan Menggunakan Metode Multiscale Retinex dan Median Filter

Abstract

The image is a representation of the light intensity function in a discrete form on a two-dimensional plane. An image is a form of information other than text, sound, and video. However, sometimes the image can also experience a decrease, namely degradation or loss of quality caused by noise, too contrasting colors, and blurring. So this study aims to improve the quality of a digital image by analyzing the use of Multiscale Retinex and Median Filter methods in improving digital images on screenshots, as well as to implement these methods in improving digital images. The data used in this study using library research, while the data used are secondary data from document studies and from library research. The method used in the improvement of digital images is the Multiscale Retinex and Median Filter methods. From the results of the application of the Multiscale Retinex method and the Median Filter on the screenshot image that has been done, it can improve the quality of a digital image. The Multiscale Retinex method can brighten digital images and the Median Filter can reduce noise/noise in digital images which makes digital image quality better. The results of several image data that have been tested are then analyzed with a histogram, the original image analyzed with a histogram shows a graph of the average pixel intensity value approaching the value 0, but after being processed using the Multiscale Retinex and Median Filter method, the average pixel intensity graph value is shown by the histogram close to the value of 250 which shows the brightness and sharpness of the digital image for the better.