Content-driven surveillance image storage optimization apparatus and method of operation

• US 20170048556A1
• Filed: 10/29/2016
• Published: 02/16/2017
• Est. Priority Date: 03/07/2014
• Status: Active Grant

• Current Assignee: Eagle Eye Networks, Inc.
• Original Assignee: Eagle Eye Networks, Inc.
• Inventors: Drako, Dean, Roskowski, Steven
• Granted Patent: US 10,412,420 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: H04N 19/91 Entropy coding, e.g. variable length coding [VLC] or arithmetic coding
• CPC Class Codes: H04N 5/915 for field- or frame-skip recording or reproducing
• CPC Class Codes: H04N 7/181 for receiving images from a plurality of remote sources
• CPC Class Codes: H04N 7/188 Capturing isolated or intermittent images triggered by the occurrence of a predetermined event, e.g. an object reaching a predetermined position signal generation from motion picture films H04N5/253

Abstract

A content-driven video storage optimization apparatus applies motion analysis at purge time or when the images are initially archived. Retention meta data is determined for image segments with and segments without motion. When triggered by a directive, segments that do not contain motion are purged. A lossy compression method optimizes storage for a security surveillance network. An appliance transforms image files into a key frame and at least one subsequent frame compatible with JPEG. A circuit filters for motion and stores metadata for each train of a key frame and its subsequent frames. The transformation operates on coefficients of frequency bins. Meta data enables content based purging of least consequential files.

Claims

1. A content-driven surveillance storage optimization apparatus comprising:

a purging circuit;

a surveillance image transformation circuit;

a surveillance image store communicatively coupled to the purging circuit and to the surveillance image transformation circuit;

a retention meta data store communicatively coupled to the purging circuit and to the surveillance image transformation circuit; and

a triggers, inhibitors, policies controller communicatively coupled to the purging circuit, whereby transformed content of the surveillance images provides retention meta data to be used by the purging circuit when activated by the controller to select images for deletion.

2. The apparatus of claim 1 wherein the surveillance image store is directly coupled to at least one surveillance image capture unit.

3. The apparatus of claim 1 wherein the surveillance image transformation circuit is directly coupled to at least one surveillance image capture unit.

4. The apparatus of claim 1 wherein the surveillance image transformation circuit comprises:

an image frame receiver which receives JPEG compatible files from a plurality of network attached security surveillance cameras;

the image frame receiver is coupled to

a subsequent frame store;

the subsequent frame store coupled to

a Huffman decoder;

a plurality of coefficient stores coupled to the Huffman decoder;

coupled to

a coefficient differencer circuit which subtracts the coefficients of the subsequent frame from the coefficients of the key frame;

the coefficient differencer is coupled to a DC level trigger which determines that one or more blocks of the image have substantially changed and records a metric into said retention meta data store.

5. The apparatus of claim 1 wherein the purging circuit comprises:

a circuit to read retention meta data;

a circuit to read purging directives;

a circuit to set a retention flag by applying the directives to retention meta data; and

a circuit to issue file operation commands according to the result of matching retention meta data with thresholds, triggers, and policies of the purging directives.

6. The apparatus of claim 1 wherein the controller triggers on free space and least or most recently used file I/O.

7. The apparatus of claim 1 wherein the controller triggers on calendars, policies, court orders, and statutory requirements.

8. The apparatus of claim 1 wherein said surveillance image transformation circuit converts a H.264 stream to segments and for each segment provides a plurality of JPEG frames.

9. The apparatus of claim 4 wherein said surveillance image transformation circuit further comprises a circuit to set a movement flag into a file header when a DC level trigger determines that one or more blocks of the image have substantially changed.

10. A method for surveillance image storage optimization comprising:

receiving one or more purging control directives;

retrieving a retention meta data array for each of a plurality of image segment identifiers;

masking each retrieved retention meta data array by each received purging control directive to set a retention flag for each image segment identifier; and

purging from surveillance image store, the image segments corresponding to image segment identifiers not set for retention.

11. The method of claim 10 wherein a retention meta data array includes a metric of relative movement within the image segment.

12. The method of claim 10 further comprising:

receiving surveillance image frames from capture devices;

storing captured surveillance image frames into a surveillance image store;

transforming high definition video into a plurality of image frames;

transforming stored image frames into image segments and retention meta data arrays; and

storing image segment identifiers and retention meta data arrays into a retention meta data store.

13. The method of claim 10 further comprising:

receiving surveillance image frames from capture devices;

transforming captured surveillance image frames into image segments and retention meta data arrays;

storing transformed image segments into surveillance image store;

transforming image frames into a series of JPEG frames; and

referencing each series of JPEG frames to an image segment.

14. The method of claim 13 further comprising:

receiving a key JPEG frame (Kframe) and a subsequent frame (Jframe);

reversing Huffman encoding of coefficients of pixel blocks;

differencing the low frequency coefficients between corresponding pixel blocks in Kframes and Jframes;

determining a metric of relative movement by normalizing the difference of low frequency coefficients; and

,

storing into retention meta data store an image segment identifier and an associated metric of relative movement.

15. The method of claim 14, further comprising:

storing a movement flag into file headers of stream segments and Kframes when a metric of relative movement exceeds a threshold.

16. The method of claim 10 wherein a retention meta data array includes a timestamp of file creation and records of file system operations on the image segment.

17. The method of claim 10 wherein a retention meta data array includes location and ownership of a capture device associated with an image segment.

18. The method of claim 10 wherein a purging control directive is inoperative on a file having a movement flag in its header.

19. The method of claim 10 wherein purging control directives include an inequality of at least one of elapsed time since capture, most recent file access, and least recent file access.

20. The method of claim 10 wherein purging control directives include ownership, location, and time of capture as specified in judicial orders, legislative statutes, and administrative policies.