High-Definition Video for Mainstream Video Surveillance

Cameras always play a crucial role in video surveillance. The latest high-definition network camera further expands the quality and reliability of video footage, but only requires low-bandwidth network capacity. Oliver Vellacott, CEO at IndigoVision, talks about the benefits and architecture of such cameras.


Over the last 10 years, video surveillance has seen a major shift from analog to digital networking solutions. In the next few years, the adoption of high-definition (HD) network video into mainstream surveillance systems will be witnessed. For HD network cameras to be widely adopted, high data rates and storage needs have to be solved.


HD network cameras have several advantages over traditional analog ones. With the whole scene recorded for later analysis, HD cameras can digitally zoom in quicker with forensic detail. They are ideal for applications where identification and recording of faces, license plates or objects are required. A single HD camera can replace several standard 4CIF cameras, thereby reducing overall cost. Retail, banking, gaming, parking, building entrances, military installations and city surveillance are typical applications.


Prices for higher resolution lenses required by HD cameras are still high. CMOS technology has to be improved quickly to overcome the low-light performance of megapixel cameras. Advanced video compression with low-bandwidth requirements must also be adopted for HD cameras to be deployed widely in mainstream surveillance applications. All of these are starting to happen, luckily.


Selection Guidelines
H.264 is the latest video codec standard, which offers higher video quality and lower compression rates. For HD cameras to be viable in mainstream applications, H.264 compression technology needs to be deployed. However, not all implementations of the H.264 standard deliver the same quality of compression.


Even when cameras claim to have H.264 implemented, data rates and storage requirements from different manufacturers' cameras vary significantly. The huge disparity in camera performance makes a significant difference in the cost of an HD video surveillance solution. Cameras with data rates of less than one megabit per second only require standard network and storage. Such cameras are also more cost-effective.


Designers and end users must know the exact data rates and storage requirements for particular HD network cameras in order to fully understand the performance and costs. These figures are often hidden and difficult to determine. A typical datasheet for a 1.3-megapixel camera from a mainstream manufacturer usually has no clear indication of data rates and storage requirements.


Some manufacturers are forced to use local storage because the bandwidth of their HD cameras is too high for live streaming across the network, negating the distributed and scalable benefits of IP video.


Architecture with Low Bandwidth
Ability to develop low-bandwidth HD network cameras is key to the adoption of HD video into mainstream surveillance. The camera must be truly IP-based with H.264 compression standard and dedicated hardware architecture. A true network camera completely eliminates any analog signal by connecting the DSP directly to the compressor chip, ensuring a noise-free picture.


Three common compression standards for HD network cameras are M-JPEG, MPEG-4 and H.264. Two types of frames used in these compression standards are the I frame, which contains the whole image, and the P frame, which only contains the information different from the previous frame. MPEG-4 and H.264 use a combination of both I and P frames, requiring less bandwidth than M-JPEG. H.264, however, only requires up to 50 percent less bandwidth than MPEG-4 to transmit the same quality image.


The H.264 standard specifies a set of optional tools which can be used to compress video. A compliant decoder must implement every tool, whereas a compliant encoder can choose which tools to use. This means that encoders from some suppliers might compress better than others.


To determine what information is transmitted in a P frame, the image has to be searched for motion in each frame. The quality of the compression depends on how far and how well the search is completed on each frame. The limitation to this searching is the available processing power in the camera, even more so at HD resolution with full frame rate.


A poor encoder design could result in higher bandwidth for good quality video, increased bandwidth during high motion, dropped frames, and blocky or blurry video. A low-bandwidth HD network camera using H.264 demands huge processing power, so it is essential that the compression engine is implemented in dedicated hardware such as FPGAs. Using HD megapixel network cameras for video surveillance monitoring offers real benefits. However, for HD cameras to move on from their current use in primarily specialist applications to everyday use requires them to have the best compression technology and low data rates.