Published Paper Details:

Our lives continue converging with an advancing innovation, which keeps progressing and changing our society. The recent development of 3d innovation brings about highly realistic and immersive resolution visuals as the resultant stereoscopic image provides a perception of the depth. The primary motivation behind the 2D-to-3D image transformation is to create the second view picture of an input 2D image that feels intentional. Conversion includes two procedures to be specific namely, the depth estimation of the given input image and the depth image-based rendering. The depth estimation from an image is a very computationally intensive process and a more challenging procedure. Depth estimation methods are been formulated into conventional and semi-automatic approaches. Conventional methods, where no human intervention is needed and an automatic computer algorithm estimates the depth for a single image, however, has not accomplished the level of quality. Therefore to accomplish a superior quality stereoscopic image, a hard segmentation algorithm is been proposed for calculating a better depth map that is fast and strong in preserving the edges. Furthermore, Image noise removal is done by using median filtering technique and the cross-bilateral filtering is used which smoothes away the local variations without affecting the strong edges which are especially reasonable for the applications with a 3d stereoscopic display.