公开(公告)号：US11935285B1

公开(公告)日：2024-03-19

申请号：US15995909

申请日：2018-06-01

Applicant: Apple Inc.

Inventor： Frederic Cao ,  Philip G. Lee

IPC: G06V20/10 ,  G06T5/20 ,  G06T5/50 ,  G06T5/00

CPC classification number: G06V20/10 ,  G06T5/20 ,  G06T5/50 ,  G06T5/73 ,  G06T5/92 ,  G06T2207/10024 ,  G06T2207/20201 ,  G06T2207/20208

Abstract: This disclosure relates to techniques for synthesizing out of focus highlighting effects in images in real-time. Digital single-lens reflex (DSLR) cameras and other cameras having wide aperture lenses typically capture images with a shallow depth of field (SDOF). So-called SDOF photography may be particularly fit for portrait photography, since it nicely emphasizes the subject, while essentially deemphasizing the background via blurring. Simulating this kind of blurring, particularly around background light sources, using a comparatively larger depth of field (LDOF) camera, may require a high dynamic range (HDR) image capture system, to accurately determine the color of background light sources, which are prone to saturation. However, small form-factor cameras having smaller lens apertures may not have the bandwidth or processing power to process full HDR images in real time. Thus, described herein are techniques to synthesize color accurate, out-of-focus background highlight regions in real-time for images captured by LDOF cameras.

公开(公告)号：US10897558B1

公开(公告)日：2021-01-19

申请号：US16566017

申请日：2019-09-10

Applicant: Apple Inc.

Inventor： Wei Sun ,  Philip G. Lee ,  Gregory Guyomarc'h ,  Frederic Cao

IPC: H04N5/222 ,  H04N5/262 ,  G06T7/194 ,  G06T5/00 ,  G06T5/20

Abstract: This disclosure relates to the lightweight and efficient synthesis of shallow depth of field (SDOF) renderings. According to some embodiments, coarse focus information and semantic segmentation information may be leveraged to generate SDOF renderings in a “live preview” or “streaming” mode. Semantic segmentation may be defined as a process of creating a mask over an image, wherein pixels are segmented into a predefined set of semantic classes. Segmentations may include as many classes as are desired by a given implementation (e.g., a ‘foreground’ class and a ‘background’ class). In some embodiments, the rendering of synthetic SDOF images according to the techniques described herein may be completed using only a single camera of a device and without the use of dedicated depth-sensing technology (e.g., structured light cameras, stereo cameras, time-of-flight cameras, etc.), thereby allowing the SDOF rendering process to operate in a fashion that is not unduly time-, processing-, and/or power-intensive.