Image Segmentation Techniques

Humans are able to easily identify the portion of the image to be segmented or separated. Computers are not that intelligent to recognize the object to be segmented accurately. For easy recognition of the object, various segmentation methods have been identified with which it becomes easy for the computer to identify the object with minimal user input. Grabcut an image segmentation technique is compared using several parameters with other image segmentation techniques. Grabcut needs minimal user interaction which is done by placing a rectangle around the object of interest. Lazy Snapping is an image cut out tool and uses Graph cut method. Intelligent scissors uses gesture motion to segment an object. Geodesic Framework scribble on the image and marks it as foreground and bcackground which is very helpful in the object segmentation. Graph cut method segments an object by drawing a cut on the graph got from the image in which the source corresponds to foreground and sink corresponds to background. All these various techniques are surveyed and compared in this paper.