The use of images has increased over the previous decade and they have the potential to be effective communication tools, similar to social media. In this technological era, uploading information or visual images to the social media seems to be gaining popularity lately. Therefore, a good image is important to provide the right information. However, the information in the image can be lost or corrupted due to the appearance of noise caused by the digitization, transmission or acquisition process. Thus, it is necessary to remove the noise before using the image for subsequent task. In this study, a new method for removing salt and pepper noise in digital image is proposed by using singular value decomposition and alpha trimmed mean median approach. The singular value decomposition will be used in the detection process by considering the distribution of pixel values in the processed image. Next, the detected noisy pixels will be replaced with a new value obtained from the trimmed alpha mean median approach. The experimental process was performed on a grayscale image with a resolution of 512×512 prepared with a salt and pepper noise density varying between 10% to 90% in order to compare the proposed method to other existing methods. The experimental results show that the proposed method has successfully reduced salt and pepper noise in high noise density. In addition, the proposed method provides better filtering results in terms of visual effects and quantitative measurement results compared to the several existing methods.
