Back to results
US 2022/0070432 A1Published application
Timing mechanism to derive non-contaminated video stream using rgb-ir sensor with structured light
AssigneeAmbarella International LP
InventorsZhe Zhang, Zhikan Yang, Ruian Xu
Life of this patent
Estimating the term…
Abstract
An apparatus includes an RGB-IR image sensor, a structured light projector, and a control circuit. The control circuit may be configured to control a shutter exposure time of the RGB-IR image sensor and a turn on time of the structured light projector to obtain a sequence of images captured by the RGB-IR image sensor, wherein the sequence of images comprises at least one image including a structured light pattern and at least one image where the structured light pattern is absent.
Claims18
An apparatus comprising:
- an RGB-IR image sensor;
- a structured light projector;
- and
- a control circuit configured to control a shutter exposure time of said RGB-IR image sensor and a turn on time of said structured light projector to obtain a sequence of images captured by said RGB-IR image sensor, wherein (i) said sequence of images comprises at least one image including a structured light pattern and at least one image where said structured light pattern is absent, (ii) said control circuit is further configured to perform a depth analysis to generate depth information using the at least one image including the structured light pattern, and (iii) said control circuit stores and executes an artificial neural network trained to (a) discern whether a face is a real face or a fake face, and (b) make a liveness determination utilizing the depth information.
1 more element
A method of implementing a low-cost structured light based 3D sensing system comprising:
- setting an exposure time of an RGB-IR image sensor based on a predetermined value;
- controlling an on-time of a structured light projector;
- capturing a sequence of images using said RGB-IR image sensor, wherein said sequence of images comprises at least one image including a structured light pattern and at least one image where said structured light pattern is absent;
- performing a depth analysis to generate depth information using the at least one image including the structured light pattern;
- and
- distinguishing whether a face is a real face or a fake face, and making a liveness determination utilizing an artificial neural network applied to the depth information.
3 more elements
Dependencies are inferred from the claim text.