Browse Prior Art Database

Method and Apparatus for motion detection based video compression, recording and surveillance Disclosure Number: IPCOM000126024D
Original Publication Date: 2005-Jun-28
Included in the Prior Art Database: 2005-Jun-28
Document File: 4 page(s) / 87K

Publishing Venue



The key invention is the architecture of the motion analysis that separated the foreground object and the background pixels, thus drastically reduce the amount of information that needed to be recorded.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 4


AUS820041929 Derek S Jennings/Watson/IBM

Oliver K Ban

Method and Apparatus for motion detection based video compression, recording and surveillance


The current video surveillance systems are normally passive recording based system, typically either analog or digital based, as shown in Diagram 1:

Diagram 1: Typical digital surveillance system architecture

The recordings are time based, some time mixed with place and other text information. These time based systems are using time stamp as reference point with scanning and panning function control and associated text information recorded on the tape either with analog formatted or digital formatted. The system itself is an open ended architecture, thus the contents of the recording need to be viewed or analyzed by human, with very time consuming fashion. These reasons represented limited algorithm for images compression.

A detailed look of most picture frames we will find the following characteristics:

Optical + tape di


Text Time

Digital surveillance system


[This page contains 1 picture or other non-text object]

Page 2 of 4

Static background such as fl b d

Dynamic moving but flexible shape objects such as bees

Picture frame

Diagram 2: Entropy value of the statistically calculated motion pictures
1. Most of the pictures has big portion of the "background" that is relatively static, like flower bed.
2. There are sometimes a relatively small but relatively flexible shaped "foreground" objects that moves inside the picture frames.
3. So statistically, we can safely say that the entropy of the picture, from the inter-frame point of view, is not evenly distributed.

Portion of the solution to the above problem has formed the background of this patent.

[Solution Summary]:

The solution lay on the idea of separating the "background" pixel group and the "foreground" pixel group, inside side of the recording system, as shown in diagram 3.

Using the nature characteristics between neighbor frames thus by selectively recording combined with time stamp, a new algorithm is created:

1. The recording is not done until a motion is detected and analyzed.

2. The background frame (B frame) is stored in the separated background frame buffer as compari...