by Subhayan Mukherjee
Stereo images have been captured primarily for 3D reconstruction in the past. However, the depth information acquired from stereo can also be used along with saliency to highlight certain objects in a scene. This approach can be used to make still images more interesting to look at, and highlight objects of interest in the scene. We introduce this novel direction here. Even though we use depth from stereo in this project, our approach is applicable to depth data acquired from any sensor modality.
The above diagram shows an overview of our proposed method. Depth and saliency information from the source image are fused and thresholded to identify the salient regions of source image in 3D. The nearest cluster of salient regions are designated as the Regions of Interest (ROI), and kept intact. Depth of Field (DoF) effect is rendered by increasing defocus (Gaussian blur) with increasing distance from the ROI, as shown in the below set of images. Thus, the proposed method automatically selects the ROI, whereas the rival methods require manual ROI selection by the user to render DoF.