Filterisasi Noise Pada Citra Uang Logam Indonesia

Abstract

Images are familiar to the computer sciences and application systems. Image or another term for the image is one multimedia component that plays an important role as a form of visual information The image consists of a set of data that contains a lot of information. Images that contain a lot of information can sometimes experience quality degradation as there is noise, color is too contrast or blur, and others. In this study will be compared the value of MSE and PSNR filter coin image 500 given noise. There are 3 sounds attached to the image of 500 coins namely Exponential, Gaussian, and Salt and Pepper noises. After coin image 500 is given noise, filtering process is done. The filter used in this study there are 3 types of filters are Average, Median and Wiener. The value of MSE and PSNR is obtained from the comparison between the image of the filtering result and the image before it is given noise. Testing is done by comparing the values of MSE and PSNR on each given image and then searching for the best filter among the mean, Median, and Wiener filters. Test results showed that the Wiener filter had the lowest MSE value and the highest PSNR for each filtered noise. Based on the results of the Wiener filter test is the best filter that can be used for the filtering process of Gaussian Exponential noise and Salt and Pepper compared with the average filter and Median.