Meeting the Megapixel Challenge for Video Security
Posted on Fri, Feb 25, 2011
Submitted by Pelco and Schneider Electric, Gold Sponsor at the ASG Security Summit & Expo
Megapixel video security imaging offers the promise of amazingly detailed video information, which can aid in forensic investigations, provide better situational awareness, and guide proper decision making in response to real-time events. However, megapixel video security imaging presents technical challenges, which must be addressed by vendors, integrators and IT administrators. Specifically, megapixel video creates challenges for low light performance and bandwidth consumption.
Let’s consider low light performance first. For a given image format size, CMOS sensors are typically megapixel (MPx) arrays (as compared to standard definition CCDs), which results in less light gathered in each pixel. Active circuits in CMOS share light gathering space, which results in less light gathered in each pixel. Also, in the past, the smaller pixels in CMOS imagers yielded an effectively lower Signal-to-Noise Ratio, when compared with CCD imagers. You can see the issues this could cause when deploying MPx cameras in low-light environments. The good news is the advancements in CMOS technology have eradicated the low-light performance gap between CCD and CMOS, making CMOS a viable or even preferable choice for low-light imaging.
Now to bandwidth consumption, storage and image file sizes. More pixels mean more data, which means larger files for each image captured by an MPx camera. When you consider that a typical 1.3 MPx camera consumes 2 to 5 MB/s of bandwidth (or higher!), it’s easy to see how you can quickly use up all the bandwidth in the network while filling up all available storage. H.264 encoding offers the ability to optimize image quality while minimizing file size, particularly compared to previous compression schemes (MJPEG, earlier parts of MPEG-4, etc.). It’s important for integrators and system designers to understand that not all flavors of H.264 are the same. Some manufacturers only utilize Base Profile H.264, which offers some improvement in compression, but not as much as High Profile H.264, which offers the greatest benefits and flexibility in compression and image quality. Integrators should also be aware that the effectiveness of H.264 compression varies, depending upon the complexity of the surveillance scene – scenes with more motion and more varied lighting will create larger image sizes. However, even though the incremental gain in compression (from Base profile H.264 to High Profile H.264), may seem relatively small on an individual camera scale, when you start to multiply it across tens or hundreds of cameras – that additional bandwidth savings can really add up.
The type of storage system also has a big impact on overall system cost and design. Traditional COTS data servers, while being relatively inexpensive, are designed for the “write once, read many” duty cycle of traditional data applications. With video surveillance, the burden placed on storage is opposite – because video is being recorded all the time, so storage devices need to be able to write data to hard disks all the time. If the storage sub-system is not tuned for massive amounts of write operations, the cost of deploying a storage solution increases dramatically. Some manufacturers have designed and developed video storage hardware specifically for the special challenges of recording video surveillance data. This should be an important consideration in any video system design decisions.
Summary:
Matching cameras and recording/storage using High profile H.264 
will give users the greatest outcome, allowing them to maximize the benefits of HD/Megapixel imaging while mitigating or minimizing bandwidth issues on their network.
How would you use this information in your planning in the future?