|
Products
|  |
| Human Dynamics Surveillance Module |
Recent advancements in computing power have enabled more complex code computation with faster real time processing, which ultimately creates a comfortable environment for artificial intelligence codes to tap on to. Up till recently, DVR (Digital Video Recorder) or Digital Security Surveillance (DSS) systems were only able to perform video data compression and decompression onto high input output (I/O) components due to restrictions in processing power. However, with increased computing power, this is no longer an issue for processors.
|
|
|
Conventional DSS/DVR which performs only recording and motion detection in digital format will soon be replaced by artificial intelligence systems which are able to analyze moving objects against backdrops with enhancements that monitor human behavior and object timing. Movements that might not be captured by the human eye can now easily be detected through these systems. This new technology in surveillance is known as the Human Dynamics Surveillance (HDS) Module. HDS Module types are varied and can range from the "Gang Fights Module" to the "Child - Maid Abuse Module." We predict that tracking objects/persons will become part of the standard surveillance modules used to aide surveillance monitoring on CCTVs at mass transportation terminals.
Intel® Digital Security Surveillance with the HDS Module will integrate artificial intelligence tracking algorithms from Independent Software Vendors, onto codes optimized by Embedded Intel® Architecture platforms. These will provide the ultimate performance systems running on SSE3, Hyper-Threading Technology and next generation Intel® 64 Architecture.
The HDS Module will have 2 key components: the first is the "intensive" algorithm which is an algorithm provided by Intel's Open-Source Probabilistic Networks Library (PNL) to track objects; the second is a "rules and criteria" algorithm, used to analyze objects identified by the object tracking algorithm to determine different human behavior which will then trigger the appropriate alarm. |
|
|
 |
|
Part of the algorithm source codes are available for download below: |
|
|
|  | |
|
