Pengaruh Estimasi Nilai Varians Pada Citra Terdegradasi Noise Spickle

Abstract

Image processing is a multimedia application. Image processing is a process to obtain image quality. Measurement of image quality can be done subjectively and objectively. However, the quality measurement that is widely used is the objective measurement of this method by preparing the original image with the distorted image which is then compared. Measurements were carried out statistically using the Mean Square Error (MSE) method. Noise is a disturbance caused by digital data storage received by the image data receiver which interferes with image quality. This disturbance is usually caused by optical physical disturbances in the image capture device. One of the noise contained in the image is noise speckle. This noise is marked by giving a black mark on the part in the form of a point that is degraded by this noise. The purpose of this research is to observe the effect of changes in the value of variance given to the original image degraded by noise speckle and its effect on the MSE value. The process is done by generating Speckle noise to be degraded to the original image by assigning variable variance values and determining the obtained MSE value.

Based on the observations, it was found that with the increase in the value of the variance of the noise speckle, the degradation of the noise also increased because the object in the image could not be recognized due to the increase in grain-like structure. The 100% noise case means that every value of the pure intensity from 0 to 225 will have the same value. The value of the variance given from 0.05 to 0.50 is also obtained that the MSE value increases.