A New Hybrid Filter Approach for Image Processing

Abstract

Today, with the rapidly advancing technology, the importance of image processing techniques is increasing. Image processing is used in many areas from facial recognition to plant disease identification. One of the important image processing stages is the filtering stage used for smoothing images and object detection. Among these filtering techniques, basic filtering techniques such as mean, median and Gaussian are used in image processing. However, these filtering techniques are known to be insufficient to achieve the desired results in some cases. In this study, a new hybrid filtering approach named Mean-Median-Gaussian (MMG) is presented using these three basic filtering techniques. It has been demonstrated that the obtained MMG hybrid algorithm gives more successful results than these three basic filtering techniques in smoothing the images and determining the boundary lines.