A method of image processing is provided for separating an image object from a captured or provided image according to a three-dimensional (3D) depth and generating a synthesized image from the image portions identified and selectively modified in the process. The method retrieves or determines a corresponding three-dimensional (3D) depth for each portion of an image, and enables capturing a selective portion of the image as an image object according to the 3D depth of each portion of the image, so as to synthesize the image object with other image objects by selective processing and superimposing of the image objects to provide synthesized imagery.