SteReFo: Efficient Image Refocusing with Stereo Vision

Benjamin Busam, Matthieu Hog, Steven McDonagh, Gregory Slabaugh
2019 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)  
Whether to attract viewer attention to a particular object, give the impression of depth or simply reproduce humanlike scene perception, shallow depth of field images are used extensively by professional and amateur photographers alike. To this end, high quality optical systems are used in DSLR cameras to focus on a specific depth plane while producing visually pleasing bokeh. We propose a physically motivated pipeline to mimic this effect from all-in-focus stereo images, typically retrieved by
more » ... mobile cameras. It is capable to change the focal plane a posteriori at 76 FPS on KITTI [13] images to enable realtime applications. As our portmanteau suggests, SteReFo interrelates stereo-based depth estimation and refocusing efficiently. In contrast to other approaches, our pipeline is simultaneously fully differentiable, physically motivated, and agnostic to scene content. It also enables computational video focus tracking for moving objects in addition to refocusing of static images. We evaluate our approach on publicly available datasets [13, 33, 9] and quantify the quality of architectural changes. Contributions and Outline. We present a general approach that utilizes stereo vision to refocus images and videos (cf. Fig. 1 ). Our pipeline, entitled SteReFo, leverages the state-of-the-art in efficient stereo depth estimation to obtain a high-quality disparity map and uses a fast, dif-
doi:10.1109/iccvw.2019.00411 dblp:conf/iccvw/BusamHMS19 fatcat:7oxfmmmgd5afjaaq4aztngcbye