23‐4: Invited Paper: Depth from Asymmetric Defocus using Color‐Filtered Aperture

Abstract

a) Prototype camera lens with CFA (b) Captured image (c) Depth map (d) Restored image Figure 1.Overview of the proposed method. We propose a novel single-shot depth acquisition method with a camera lens in which a color-filtered aperture (CFA) inserted. The captured image (b) taken with the prototype CFA (a). The CFA generates asymmetric defocus blurs. These asymmetric defocus blurs are confirmed as colored blurs in a region outside the focal plane in the captured image (b). The close-up image in (b) is captured in front of the focal plane and has the colored blurs, for example, the blurred edge of a white object is slightly colored with blue (left) and orange (right). By utilizing the blurs, the depth map (c) is estimated. The depth map (c) and the restored image (d) without the colored blurs are generated simultaneously because the depth estimation is equivalent to eliminating the colored blurs. AbstractWe propose a new method, Depth from Asymmetric Defocus using a Color-Filtered Aperture inserted into the camera lens aperture. The proposed method provides a highly accurate depth map from a single input image by combining a depth from defocus approach with color-filtered aperture photography. Author Keywords3D image sensing; depth from defocus; computational photography.

Physical Cue based Depth-Sensing by Color Coding with Deaberration Network

“…Instead of utilizing contextual information, color-coded aperture (CCA) methods can acquire a depth map based on a physical depth cue encoded in a single-shot image by inserting different types of color filters [6,8,20,24,28,29,30] into the lens aperture as shown in Figure 1: (a)Our prototype camera with color coded aperture and its captured image. (b)Depth map estimated by a stereo camera for the reference.…”

“…(b)Depth map estimated by a stereo camera for the reference. (c)Depth from the analytical defocus (DfAD) [29] (d)Depth obtained by the proposed method with a deep deaberration network (DDN). In the distant regions, our method is superior to the conventional one.…”

“…Furthermore, the depth cues often disappear because of several uncertainties, such as saturation, soft shadow, dark color and large blur. These uncertainties are not treated distinctly in conventional methods [6,8,15,20,24,28,29,30].…”

“…CCA methods [6,8,20,24,28,29,30] are categorized to a computational photography (CP) technique [39] developed in the last decade. The image quality of CCA is higher than codedaperture [25] having unnatural blur shape due to the special shape of the aperture.…”

“…The image quality of CCA is higher than codedaperture [25] having unnatural blur shape due to the special shape of the aperture. In order to acquire the depth map, CCAs use disparity [6,20,24,30] or defocus blur [8,28,29].…”