Method and system for non-causal zone search in video monitoring

Patent metadata

Field	Value
Publication number	US-9779307-B2
Application number	US-201414510029-A
Country	US
Kind code	B2
Filing date	Oct 8, 2014
Priority date	Jul 7, 2014
Publication date	Oct 3, 2017
Grant date	Oct 3, 2017

Abstract

A computing system processes a video recording to identify a plurality of motion events, each corresponding to a respective video segment along a timeline of the video recording. The computing system identifies at least one object in motion within a scene depicted in the video recording and stores a respective event mask for each event. The computing system receives a definition of a zone of interest within the scene. In response to receiving the definition, the computing system determines, for each motion event, whether the respective event mask of the motion event overlaps with the zone of interest by at least a predetermined overlap factor, and identifies one or more events of interest from the plurality of motion events, wherein the respective event mask of each identified event of interest is determined to overlap with the zone of interest by at least the predetermined overlap factor.

First claim

What is claimed is:

1. A method of facilitating review of a video recording by performing a retrospective event search, comprising:
processing the video recording to identify a plurality of motion events, each motion event corresponding to a respective video segment along a timeline of the video recording, and identifying at least one object in motion within a scene depicted in the video recording;
obtaining and storing a respective event mask for each of the plurality of motion events identified in the video recording, the respective event mask including an aggregate of motion pixels from multiple frames of the motion event, wherein the motion pixels are associated with the at least one object in motion in the multiple frames of the motion event;
after processing the video recording, receiving a definition of a zone of interest within the scene depicted in the video recording; and
in response to receiving the definition of the zone of interest:
determining, for each of the plurality of motion events, whether the respective event mask of the motion event overlaps with the zone of interest by at least a predetermined overlap factor; and
identifying one or more events of interest from the plurality of motion events, wherein the respective event mask of each of the identified events of interest is determined to overlap with the zone of interest by at least the predetermined overlap factor.

2. The method of claim 1 , wherein obtaining and storing a respective event mask for each of the plurality of motion events identified in the video recording comprises:
generating the respective event mask for each of the plurality of motion events, wherein the generating includes:
creating a respective binary motion pixel map for each frame of the respective video segment associated with the motion event; and
combining the respective binary motion pixel maps of all frames of the respective video segment to generate the respective event mask for the motion event.

3. The method of claim 1 , further comprising:
receiving a first selection input from the user to select the zone of interest as a first event filter; and
visually labeling the one or more identified events of interest with a respective indicator associated with the zone of interest in an event review interface.

4. The method of claim 1 , wherein the definition of the zone of interest includes a plurality of vertices specified in the scene of the video recording.

5. The method of claim 1 , further comprising:
after receiving the definition of the zone of interest:
processing a live video stream depicting the scene of the video recording to detect a start of a live motion event;
generating a live event mask based on respective motion pixels associated with a respective object in motion identified in the live motion event;
determining, in real-time, whether the live event mask overlaps with the zone of interest by at least the predetermined overlap factor; and
in accordance with a determination that the live event mask overlaps with the zone of interest by at least the predetermined overlap factor, generating a real-time event alert for the zone of interest.

6. The method of claim 5 , further comprising:
visually labeling the live motion event with an indicator associated with the zone of interest in an event review interface.

7. A computing system for facilitating review of a video recording by performing a retrospective event search, comprising:
one or more processors; and
memory having instructions stored thereon, the instructions, when executed by the one or more processors, cause the processors to perform operations comprising:
processing the video recording to identify a plurality of motion events, each motion event corresponding to a respective video segment along a timeline of the video recording, and identifying at least one object in motion within a scene depicted in the video recording;
obtaining and storing a respective event mask for each of the plurality of motion events identified in the video recording, the respective event mask including an aggregate of motion pixels from multiple frames of the motion event, wherein the motion pixels are associated with the at least one object in motion in the multiple frames of the motion event;
after processing the video recording, receiving a definition of a zone of interest within the scene depicted in the video recording; and
in response to receiving the definition of the zone of interest:
determining, for each of the plurality of motion events, whether the respective event mask of the motion event overlaps with the zone of interest by at least a predetermined overlap factor; and
identifying one or more events of interest from the plurality of motion events, wherein the respective event mask of each of the identified events of interest is determined to overlap with the zone of interest by at least the predetermined overlap factor.

8. The computing system of claim 7 , wherein obtaining and storing a respective event mask for each of the plurality of motion events identified in the video recording comprise:
generating the respective event mask for each of the plurality of motion events, wherein the generating includes:
creating a respective binary motion pixel map for each frame of the respective video segment associated with the motion event; and
combining the respective binary motion pixel maps of all frames of the respective video segment to generate the respective event mask for the motion event.

9. The computing system of claim 7 , wherein the operations further comprise:
receiving a first selection input from the user to select the zone of interest as a first event filter; and
visually labeling the one or more identified events of interest with a respective indicator associated with the zone of interest in an event review interface.

10. The computing system of claim 7 , wherein the definition of the zone of interest includes a plurality of vertices specified in the scene of the video recording.

11. The computing system of claim 7 , wherein the operations further comprise:
after receiving the definition of the zone of interest:
processing a live video stream depicting the scene of the video recording to detect a start of a live motion event;
generating a live event mask based on respective motion pixels associated with a respective object in motion identified in the live motion event;
determining, in real-time, whether the live event mask overlaps with the zone of interest by at least the predetermined overlap factor; and
in accordance with a determination that the live event mask overlaps with the zone of interest by at least the predetermined overlap factor, generating a real-time event alert for the zone of interest.

12. The computing system of claim 11 , wherein the operations further comprise:
visually labeling the live motion event with an indicator associated with the zone of interest in an event review interface.

13. A non-transitory computer-readable medium for facilitating review of a video recording by performing a retrospective event search, the non-transitory computer-readable medium having instructions stored thereon, the instructions, when executed by one or more processors, cause the processors to perform operations comprising:
processing the video recording to identify a plurality of motion events, each motion event corresponding to a respective video segment along a timeline of the video recording, and identifying at least one object in motion within a scene depicted in the video recording;
obtaining and storing a respective event mask for each of the plurality of motion events identi

Assignees

• Google Inc

Inventors

• Laska Jason N
• Nelson Gregory R
• Duffy Greg
• Mitsuji Hiro
• Hill Cameron
• Davidsson Martin
• Montalbo Michael D
• Wan Tung Yuen

Classifications

• H04L63/0428 Primary

  wherein the data content is protected, e.g. by encrypting or encapsulating the payload · CPC title

• G08B13/19604

  involving reference image or background adaptation with time to compensate for changing conditions, e.g. reference image update on detection of light level change · CPC title

• H04N23/651

  for reducing power consumption by affecting camera operations, e.g. sleep mode, hibernation mode or power off of selective parts of the camera · CPC title

• H04N23/65

  Control of camera operation in relation to power supply · CPC title

• H04N23/6811

  based on the image signal · CPC title