Gradient Method to Construct 3D Digital Image of a 2D Digital Image

Kaushal Patel

Abstract: 3D vision of image has become the area of interest of many researchers. Numerous image processing algorithms have been developed over the past decade. This work covers a novel method of converting 2D image to 3D image.3D imaging provides very detailed and valuable information. The approach effectively incorporates visually prominent boundaries, prominence measure of pixels and local color & region gestures. The proposed algorithm uses the method of gradient that would be the direction of uphill. The gradient has a direction and a magnitude, it is natural to encode this information in a vector. The length of this vector provides the magnitude of the gradient. Vector can be used to provide height/depth of an image, that generates 3D image of an original 2D digital image.

Keywords: 3D, 2D, reconstruction, gradient, displacement map,