	US 12,236,683 B2
	Image/video analysis with activity signatures
Joshua Migdal, Wayland, MA (US); and Vikram Srinivasan, North Billerica, MA (US)
Assigned to NCR Voyix Corporation, Atlanta, GA (US)
Filed by NCR Voyix Corporation, Atlanta, GA (US)
Filed on Dec. 3, 2021, as Appl. No. 17/541,716.
Application 17/541,716 is a division of application No. 16/943,140, filed on Jul. 30, 2020, granted, now 11,475,669.
Prior Publication US 2022/0092312 A1, Mar. 24, 2022
Int. Cl. G06V 20/40 (2022.01); G06F 16/71 (2019.01); G06F 16/738 (2019.01); G06F 18/24 (2023.01); G06N 20/00 (2019.01); G06T 7/269 (2017.01); G06V 10/22 (2022.01); G06V 10/72 (2022.01); H04N 7/18 (2006.01)

CPC G06V 20/47 (2022.01) [G06F 16/71 (2019.01); G06F 16/739 (2019.01); G06F 18/24 (2023.01); G06N 20/00 (2019.01); G06T 7/269 (2017.01); G06V 10/22 (2022.01); G06V 10/72 (2022.01); H04N 7/18 (2013.01); G06T 2207/10016 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30196 (2013.01); G06T 2207/30232 (2013.01); G06V 20/44 (2022.01)]

20 Claims

1. A method, comprising:

obtaining video frames from a video feed;

identifying a path of activity laid out within of certain video frames for a modeled activity;

calculating a measure over the path based on movements of an item or a person associated with the path;

aggregating the measure along the path at each point on the path as an aggregated measure;

accumulating the aggregated measure into a data structure across a plurality of additional video frames from the video feed as a time series;

normalizing the data structure for a monitored activity appearing in the video feed and associated with the modeled activity; and

compressing the video frames into the data structure, wherein the data structure comprises a series of one-dimensional stacked frames, each frame compressed into a single row of numerical pixel values, with each row stacked on previous rows to form a single two-dimensional image representative each of the corresponding frames of the video feed.