An Improved Region Based Segmentation By Using Edge Detection and Filter Technique

Abstract

In computer vision effective image segmentation and object recognition are an important task. Interactive image
segmentation is an important part in medical images, remote sensing and real time images .This paper presents an effective
interactive image segmentation based on region merging method with fuzzy switching median filter, canny Edge detection &
Anisotropic diffusion filter Technique . For natural images or Medical images, completely automatic image segmentation is
usually inflexible. Interactive schemes are very helpful with a few simple user inputs are high quality solution. Experimental
results shows that proposed scheme is reliably extract the object contour from the background and making the edges sharp.
The main objective of this present work is to provide an efficient approach for image merging using MSMR and for increase
the accuracy of the segmented image, we use fuzzy median filter and Anisotropic Diffusion filter. For edge detection, we use
canny edge detection method .