Video Intelligence
CCTV networks are rapidly increasing in size and it is not uncommon to come across systems with thousands of cameras. As a result, the number of cameras which each CCTV operator has to monitor is also increasing. This can result in operators missing crucial events. COE VI is a high-end video intelligence product that has been designed by seasoned security professionals with years of field experience. It is engineered from the ground-up to accommodate large-scale camera networks with unrivaled ease of use, surveillance performance, expandable functionality and network scalability.
COE VI - A unique video intelligence approach
Not all video intelligence suites are the same. Prior to COE VI, there were two type of analysis solutions available in the market:
ON CHIP ANALYSIS
The entire analysis takes place on the codec on uncompressed video, thus minimizing the false alarms due to artifacts caused by compressing the video. But this process contends with video compression for resources, thus degrading the quality of the compressed video. Additionally, it is only possible to have one analysis function per camera, i.e. not possible to have perimeter security and abandoned object detection using the same camera. With video intelligence technology evolving rapidly, on chip analysis also hampers future upgrades as one is limited by the DSP already installed on the device and hence can’t take advantage of future upgrades to technology
ON SERVER ANALYSIS
The video is brought back to a central control room and then analysis takes place on this compressed video. On one hand, this process allows multiple detection rules per camera and also allows future upgrades to video analysis detection but the artifacts introduced by video compression reduces the performance of the analysis, thus increasing the false alarm rate. Also, bringing back the video at high quality uses a lot of bandwidth and since each stream then needs to be decoded on the PC, it is only possible for a PC to run the analysis on
COE VI SYSTEM ARCHITECTURE
As can be seen above, there are shortcomings to both the aforementioned approaches. COE VI offers a unique hybrid architecture which offers the best of both worlds
COE VI has a unique architecture which distributes analysis between X-Net Codec G3 and a backend server. This ensures that the analysis happens on uncompressed video, yet the bulk of processing is carried out at the backend server which can support multiple detection rules per camera and can also be easily upgraded to support additional rules as and when these become available.
As a result, the system scales up extremely well, with a single server capable of handling up to 100 video feeds simultaneously and at the same time bandwidth utilized on the network is reduced since high quality video is only sent to the control room when there is an incident.
