Optical Flow


Description:
Optical Flow is a set of techniques used to determine characteristics of an object's motion in a video sequence.


Implementation:
The Optical Flow Engine is the set of image processing methods implemented as a cross-platform library.


Features:
a) Can be used for dense (for each pixel) or sparse (for particular pixels) optical flow calculation
b) Fast processing – 5 fps on a mobile device
c) Is based on state-of-art methods of calculation of optical flow and on original customizable algorithms
d) Implementations for both monocular and binocular camera systems.


Example Applications:
a) Background Subtraction
b) Motion detection and recognition
c) Classification of human motion patterns (gestures, walk style, etc.)
d) 3D reconstruction using both monocular and binocular camera systems
e) Security and alerting systems
f) Human fatigue and general health state monitoring by analysis of body motion characteristics


Adjacent Areas:
a) Augmented Reality
b) Object Tracking
c) Localization and Mapping

© 2020 All Right Reserved. Square Nest Research Labs Inc.
mp@square-nest.com

Made on
Tilda