While three-megapixel to five-megapixel cameras offer a lot of advantages, further technological improvements will be needed before users take them up in a big way. Even though multimegapixel cameras present good value propositions, problems with sensitivity, storage and display will not be solved until H.264 adoption becomes widespread.
According to IMS Research, about 80 percent of all installed CCTV systems are analog, so strong growth is predicted for megapixel network cameras. "The IMS study confirmed that the entire industry is undergoing a period of sustained growth," said Mobotix CEO Ralf Hinkel. "The market for network cameras is set to increase by an average of 40.1 percent annually between 2006 and 2011, with the market for network video surveillance software not far behind at 35.8 percent."
Paul Bodell, Chief Marketing Officer of IQinVision, estimates that North America for 45 percent of the IP video market, Europe for 45 percent and the rest of the world accounts for the remaining 10 percent. "There are very few providers when it comes to three-megapixel to five-megapixel single lens cameras," Bodell said.
More exact figures were provided by Hidenori Taguchi, Manager of the Product Planning Department at Sony Corp. "The network camera market was worth US$500 million in 2007," he said, "with the megapixel market worth $80 million." To calculate market size for three-megapixel and above cameras, he estimated that less than three percent of the network camera market is multi-megapixel; this would equate to $15 million per year. "You have to remember that roughly 90 percent of the megapixel camera market is only 1.3-megapixel cameras," he said.
"We have seen many companies try to enter this market," said Bodell. "Some have tried to market 20-megapixel cameras, but the market is just not there. These companies are redefining portfolios to offer three-megapixel to five-megapixel cameras." He also noted that some big players have been talking about entering. "We expect them to launch products in the next year to 18 months."
Applications
According to Grandeye CEO Alexa McCulloch, the strongest vertical markets for 360-degree network cameras are commercial security, liability protection, safety, enterprise business, military applications, robotics and defense.
In Europe, IQinVision is providing service for a lot of city surveillance. "This is just starting to grow in North America," said Bodell, while pointing out that the education vertical is very strong in the U.S. while commercial and industrial applications are strong in both regions. In addition, transportation is an increasingly important segment in both Europe and North America.
Bodell also pointed to applications outside security, including education (administrators may wish to survey campuses to see if there is snow to cancel classes), retail (video analytics such as people counting and monitoring), license plate recognition for parking lots, trash collection (systems associate vehicle with weight and amount being dumped, while recording license plate number and recording information to invoices) and hospitals (medical diagnostics to monitor vital signs). Facial recognition and left-luggage video analytics are other possibilities.
Latest Technology
The latest technology, in Bodell's view, involves single-lens day night cameras and on-camera recording. "Until just recently," added Bodell, "traditional box cameras with one megapixel to two megapixels have been most popular; but demand is starting to shift toward three-megapixel to five-megapixel cameras because of the packaging and housing that IQinVision is using."
Mobotix is moving to a decentralized concept with high-resolution cameras that have built-in intelligence, internal DVRs (SD cards), the use of the Mobotix video codec MxPEG and a new Q22 hemispheric camera with 360-degree recording and virtual PTZ with no moving parts. "Usually," said Hinkel, "cameras supply only images, while the processing and recording is done later on a central PC using video management software." "This traditional centralized structure, however, is not suitable for high-resolution video systems," he said, "since it requires not only high network bandwidth, but also enormous PC processing power to support several cameras."
"Unlike traditional systems," said Hinkel, "the decentralized Mobotix concept incorporates a high-speed computer and, if necessary, a digital memory (SD or flash card) for long-term recording in every camera. The PC is now used only for viewing, not for analysis or recording. As a result, Mobotix cameras record events even without a running PC and digitally record videos with sound for archiving purposes."
Grandeye recently introduced its Halocam compact IP series of ultra-high-resolution network cameras. "Each," said McCulloch, "features patented, award-winning 360-degree technology, 12-bit color pipeline, five-megapixel sensor and ability to generate multiple virtual cameras. Unique in-camera intelligence combined with a powerful three-dimensional graphic chips enable a variety of advanced 360-degree video analytic functions, including threat recognition, virtual trip wires, changes in field of view and automated multiperson PTZ tracking triggered by motion detection and/or external alarms."
Ultimately, the primary benefit of using high-resolution cameras is that the stored images are more revealing and provide stronger evidence than other formats. Other criteria include IP65, no moving parts, robust and low maintenance, no additional power or heating required, no software and licensing costs, sun and backlight compensation, dual camera technology: two-in-one intelligent recording technology to reduce required storage, recording of events even without a running PC and digitally record videos with sound for archiving purposes. Another attraction is lower network bandwidth because everything is processed in the camera itself, and the high-resolution images do not have to be constantly transferred for analysis.
Challenges
Several challenges remain. "We are still fighting the giants in the market," said Bodell. "These," he added, "still promote mostly analog technology and are, thus, the dominant CCTV forces. They imply that there is value to IP cameras, while still marketing analog. We are fighting this momentum.
The second challenge, in Bodellˇs view, is to reach people with straightforward messaging. "As it is a very immature market, at least from the marketing perspective, there is a tendency to confuse people with an alphabet soup of technology that does not translate value to the customer. We focus on the value of megapixel cameras rather than talking endlessly about functions and features."
For Taguchi, the biggest challenge is sensitivity. "Codecs are mostly based on JPEG, though H.264 is being developed. If you increase the number of pixels, then normally sensitivity is lowered. Since sensitivity is the most important factor in surveillance for security purposes, it has a big effect on the industry."
Sony, he said, is not yet making three-megapixel or five-megapixel cameras. "We simply have not been able to get the sensitivity with currently available sensors." Another impediment is display and storage. "How do you store and display such large images?" asked Taguchi. ¨Most displays can handle up to two megapixels (1,980-pixel by 1,080-pixel images), but anything over that is problematic." Likewise, he noted, JPEG is not a very efficient method of compression. "To record and store three-megapixel and five-megapixel images requires servers with a lot of capacity."
Oh Tee Lee, Regional Director, South Asia Pacific, Axis Communications, agreed. "H.264, a new compression standard, will have a big influence on this market, but it has to be widely accepted or there will simply not be sufficient bandwidth or storage capacity to enable use of megapixel technology." In Lee's estimation, the use of megapixel will be dramatically expanded over the next 12 months to 24 months. "Every day, in fact, there are more requests for megapixel."