Surveillance

High definition, but flexible

By Jianggan Li | 22 May 2009

After last Saturday’s corrosive-liquid attack in a busy pedestrian street in Mong Kok, Hong Kong government announced that the planned CCTV installation is going to be expedited.

View photos

Related Categories

• SURVEILLANCE

From this Section

• INTERVIEW

Thirty people were injured in the attack, which is the second such case in Mong Kok since last December. After the first attack, the local district council decided to set up ‘eyes in the sky’ – a network of surveillance cameras which they say will deter similar incidents and help bringing the culprits to justice.

A HK$400,000 (US$51,600) budget for the installation was approved in January. However the police advised the council to better install ‘high definition’ cameras which would cost HK$1.7 million (US$219,000), according to Hau Wing-cheong, the district councillor.

There is currently no standard definition of high definition cameras, and the general consensus in the industry is that megapixel surveillance cameras fall in this range.

However, just having megapixel cameras is different from high definition surveillance, explains Yoshikazu Hirano, Sony’s Asia Pacific General Manager for Security Solutions. “All the components, including the network, recording device, the storage and the display need to be compatible to make the megapixels worthwhile.”

Hirano also cautions that having a high definition system doesn’t mean replacing all the legacy cameras with high definition ones. “High definition cameras only need to be installed in areas where they can really help with what conventional cameras can’t do,” he says. “An example of this is recognising people and behaviours in business transport environment.”

While high definition images and video streams dramatically increase the pressure on storage systems, Hirano says the alternative is to make the system more flexible. “Even in a specific environment where high definition is necessary, it is not always required,” he notes. “The system needs to be flexible to switch between high definition and low definition modes to reduce the burden on storage.”

Commenting on the gigapixel cameras which start emerging, Hirano is of the view that normal applications of security need anything more than 1.3 in the next three to five years.. “There are a lot of issues with ultra-high definition cameras,” he says.

One of such issues is the lighting, as when more space is watched, more illumination would be required for the images and video to be sufficiently lit for identification of people and behaviours. And covering a wide city in most settings would create more blind spots, which would means more cameras need to just cover these areas.

Therefore it would only be useful in very specific settings, says Hirano. One example is the stadium where a wide, unobstructed area needs to be covered and the venue is sufficiently lit.

“High definition is not just about resolution, it’s also about how people use the resolution,” Hirano concludes. “A clear surveillance strategy addressing the needs in a particular setting will be much more cost effective.”