Experts talk trends in image resolution, edge capabilities and video storage at ASIS 2013.
Editor’s note: This is part two in a two-part series on video surveillance trends from the perspective of several industry experts. Part one examines overall industry trends, the continued migration to IP and how companies are continuing to provide support to the large existing base of analog device users. Part two delves into the progression of high-resolution imaging technology, developments at the edge and the future of video recording.
Over the past several years, it seems as though there has been an arms race within the video surveillance industry when it comes to high-resolution cameras. Just a few years ago, only a handful of vendors offered cameras with high-megapixel imaging capabilities. Now HD resolution is the norm and some companies offer cameras that provide users with as much as 40 megapixels of resolution.
There have also been great advancements over the last several years at the edge. Functions such as storage and analytics that could once only be performed on the server side are now being done within the cameras themselves.
Has the market settled on a sweet spot for image resolution? What kind of impact will advancements in edge capabilities and onboard storage have on camera R&D over the next several years? Here’s what several experts had to say at the ASIS show this week in Chicago.
SIW: Has the industry settled on how much resolution is needed for given applications or are we going to continue to see higher and higher megapixels in the years to come?
Fredrik Nilsson, general manager, Axis Communications: I think one of the big drivers for IP video, in general, is the resolution and the question is how much resolution do you need and what do you need to pay for that resolution? I think, for the most part, for new installations today it’s either 720p or 1080p as a requirement and sometimes optionally 3 and 5-megapixel. That is becoming more and more the bulk of the market for video installations. Higher megapixels than that is really niche installations and one of the reasons is it’s still a little expensive to store that video, but more importantly, it’s very expensive to have the appropriate lens connected to that kind of camera and lenses don’t decrease in price like computer equipment with Moore’s Law, but that is fine mechanics and optical elements that, if anything, stays in price or goes up a little bit. And, for the most part, anything over five if you go over to 10 and 20, it becomes quite expensive on the lens side, quite niche and therefore is a relatively small part of the market. Resolution is important, but the sweet spot really is 720p up to 5-megpixels and that’s really where it is today. The other trend we see for high-resolution is 360/180-degree applications, which has been out there for a while and now with the later processors, you start to get good performance out of those cameras and you can get your full frame rate if needed.
Frank De Fina, senior vice president of sales, North America, Samsung Techwin America: We will undoubtedly see megapixel cameras with higher resolution down the road. Just look at the consumer smart phone market and you’ll see new devices like the Galaxy S4 that features a 13-megapixel autofocus camera that shoots video and stills. That’s a lot of resolution for a mobile device, which serves to raise people’s quality and performance expectations across the board. The same holds true with consumer flat screen TVs as almost everything we view every day is in HD. So it’s expected that these same viewing standards apply to security professionals. Low resolution video is simply not acceptable by any standard of measure these days and we can expect to see megapixel surveillance cameras with higher resolution and even better compression solutions moving forward.
Gadi Piran, president of OnSSI: It’s safe to assume that megapixel camera suppliers will continue to develop new and improved imaging technologies that deliver greater resolution. HD images are no longer a luxury but an expectation in today’s technology environment.
Wendi Burke, director of marketing, IQinVision: The industry standard (based on sales data) is two-megapixel/HD 1080p, however, there is still a large demand for higher resolution options for general surveillance applications and we will continue to see resolutions climb as lens options and image quality advancements make these larger resolution options increasingly viable for use in identification and prosecution.
Scott Schafer, executive vice president, Arecont Vision: There is no set standard for megapixel resolution for a specific application other than perhaps in gaming where specific details like playing card and chip identification is mandated by state regulations. Otherwise, resolution is typically determined based on users’ surveillance needs and objectives. In some instances, 40-megapixel panoramic cameras may be specified to deliver the highest attainable image quality, yet a similar application at a different facility may specify an eight- or 12-megapixel panoramic camera. I expect we will continue to see increases in the resolution of megapixel cameras as new chipsets and processing technologies continue to evolve.
Dave Poulin, director business operations, security and evidence management solutions, Panasonic: While HD and megapixel resolution security cameras have been a focus for security professionals in recent years, we don’t expect the push for higher megapixels to continue its trajectory much longer within the mass market. Instead, we anticipate more focus on new camera features and advancements in processing technologies such as LDC (lens distortion compensation) that improve image quality and optimize data captured by newer IP products. These and other advances will make it possible for security professionals to record and analyze facial features, textures, symbols and other details that was not possible with previous generations of video surveillance camera technology.
George Maroussis, manager, Genetec Technology Alliance: Even though camera vendors keep offering higher and higher megapixels and resolution, it always comes at a price in 1) storage and 2) bandwidth. While the idea of “one high-definition camera can replace 10 low-res cameras" is appealing, let’s not forget that one break down and blocked image on one camera versus 10 dispersed cameras can make all the difference. We will see higher megapixels but this will be balanced or throttled based on network bandwidth, camera optical/dynamic performance ranges (i.e. WDR, low light), as well storage and retention time.
SIW: Have advancements at the edge such as increased storage capabilities changed the way the industry looks at developing camera technology now?
Nilsson: With the cost of SD cards, I actually bought one for a camera for myself - a Class 10 SD card - and it was $23. I remember when we launched (the Axis Camera Companion software), a 32-gigabyte memory at that time cost $50. It’s really moving the industry in the direction of being able to do more on the edge and do it very inexpensively. We did a calculation that if you’re using that latest kind of (memory) size that is supported by most cameras that is 64 gigabytes, and with 64 gigabytes you can get five days worth of full frame rate recording even if you do good quality 720p. If you do motion-based (recording) or you cut down to 15-frames-per-second, you’re probably going to get many more days than that, maybe a couple of weeks. But just for the sake of comparison, you can get five days full frame rate recording into the camera for a cost of $40. If you look at any recording solution, that’s kind of difficult to compete with.
De Fina: There are multiple inter-related developments at the edge and on the cloud that will continue to change industry perceptions, as well as how and what new camera technologies are developed moving forward. New cameras, for example, feature better compression solutions like H.264 that reduce storage and bandwidth requirements. And on-board analytics like motion detection and object left behind, allow cameras to be programmed to change resolutions depending on real-time events and triggers. More advanced video management systems are also providing centralized feature enhancements that can enhance otherwise conventional cameras with greatly increased functionality. This leaves open a very wide spectrum for camera development as the arguments for intelligence at the edge or on the enterprise level continues. At Samsung, we’re addressing both alternatives by offering what we believe to be the largest.
Storage is one issue as a result of several factors including better compression like H.264 which has reduced storage and bandwidth requirements, and the ability to program megapixel cameras to switch to high-resolution only when prompted by analytics like motion detection or an external trigger from a video management solution. All of these edge devices – cameras, NVRs, control software and analytics – are all evolving in tandem to deliver higher performance and ROI.
Piran: It’s a natural expectation for recording technologies to continue to evolve and deliver greater storage capacities at lower costs. In step with improvements in edge recording technology, I anticipate that camera manufacturers will also continue to improve imaging performance in terms of higher resolution. This will make it more cost-effective for users to add cameras to their systems which will further accelerate the deployment of VMS solutions like Ocularis that offers infinite scalability without penalty.
Burke: Although talked about for years, edge technology is now just taking off. Previous limits on processing power and storage capacity are rapidly diminishing while currently emerging technologies will essentially provide more bandwidth and processing power capability within the camera than is possible on the network.
Schafer: There’s no mistaking that there are tangible costs associated with increased storage and bandwidth capacity. However, server/recording technologies also continue to improve to better accommodate images produced by high-resolution and high frame rate imaging devices such as megapixel cameras. In recent years, we have already seen developments like H.264 compression and event-based programmable resolution and frame rates have alleviated many of the concerns and issues initially experienced by high performance imaging devices. The deployment of more advanced, high resolution megapixel based video surveillance systems is testimony to this trend that continues to grow on a global scale.
Poulin: Increased storage capabilities have certainly changed the way new camera technologies are developed. Built to accommodate real-time data streaming and user expansion, digital video systems are already enabling security professionals to capture data that is of both higher quality and more advanced resolution. Increased storage capabilities allow them to record and process greater volumes of video much faster than legacy tape-based video surveillance systems, using advanced data compression technologies to maximize bandwidth and deliver exceptional quality playback.
Maroussis: Edge recording was a brilliant invention and has incredibly impacted the way cameras are becoming smarter, but edge data alone is not enough to allow the camera industry to rest. Higher resolution, panamorphic, 360-degree and multi-resolution cameras with ever increasing intelligence keep getting better and better. Smart analytics and real-time functions, PTZ via remote Wi-Fi and mobile devices, with built-in alarms and increased sensitivities to alerts, will keep the camera industry developing ever-changing technology.
SIW: For years, video storage was a big challenge for end users. Is this still true today or have storage capacities and price points reached a point where it is no longer the issue that it once was?
Olivier Thierry, chief marketing officer, Pivot3: The challenge has changed and become more IT-centric as video surveillance has gone IP and high-definition. High-resolution video drives up the need for both capacity and performance. Even though price points have come down, capacities have increased. Therefore, the need for shared IP-based storage (SANs) has increased because these appliances deliver robust performance, failover options and increased retention times. Overall, the trend is for more virtualization and converged infrastructure deployment (servers and storage in same appliances) to reduce capital costs, simplify deployment and maintenance, and reduce operations costs including power and cooling.
Mike Scirica, vice president of marketing and sales, WavestoreUSA: Storage is typically the most expensive element in a commercial video security system and the price of storage continues to outpace all other technology savings. But high-definition/megapixel cameras are reducing the total number of cameras required for an installation by covering more area. It is usually more efficient to record a single high megapixel camera than several lower resolution ones. Wide area surveillance is possible and the smaller percentage of moving through a camera's field-of-view allows for more efficient recording. Overall, this streamlines recording and reduces upfront and long-term costs.
Nilsson: I think as the industry moves over to IP, more and more of that storage moves off to the common off-the-shelf type of storage devices – servers, SD cards or even small NAS storage boxes which you can also get for probably $200 with a terabyte of storage and the DVR would have a difficult time competing with that. The DVR is mostly used in very small systems today, it has very low cost, but again, the SD cards are gradually moving in there as well, so I think there is more and more hardware being done as kind of a ready-made NVR for systems that are also competing with NAS and SD cards. I think, over the years, you’re going to see even fewer DVR companies. That’s my guess.
De Fina: All system related expenses are and will continue to be an issue, as end users continue to demand better performance for less. Storage is no exception despite new technologies like H.264 compression, analytics and control software that provide multiple means of reducing bandwidth and storage requirements. Technology evolution dictates that we will see new storage solutions with increased capacities at lower price points moving forward.
Piran: Video recording solutions continue to improve as disk densities, compression techniques and the means by which cameras can be programmed to change resolutions based on external triggers from analytics and control solutions like our Ocularis VMS. But costs still play a major role in the decision making process when specifying system performance and capabilities. We can expect all of these trends to continue as larger storage capacities in the terabyte and petabyte range continue to make news.
Burke: Storage is less of an issue than it was perceived to be in the past. The industry has grown smarter regarding how to effectively and efficiently optimize video recording and storage. Combine that with increased storage capacity and lower prices and you have the first mainstream edge technology that has really gained strong adoption and practice. In the future, we will see even more use of edge storage and integration with other technologies from VMS to analytics. The possibilities are enormous.
Schafer: There is a combination of factors that have affected storage capacities and cost issues. One is the on-going improvements in server and hard disk technologies combined with technologies like H.264 compression which conserves bandwidth and storage space. The other significant development is in the area of megapixel cameras where H.264 compression and features like the ability to change frame rates and resolution based on real-time events and triggers. This enables large numbers of megapixel cameras to be deployed and programmed to shift to high frame rates and imaging resolutions when triggered by an event via a VMS, DVR or some other control software prompt. There will always be issues to contend with when looking to strike the right balance between cost and performance, but as recording and imaging technologies continue to improve, both storage and bandwidth capacities will become even less of an issue than they are today.
Poulin: While there are still high costs associated with video storage, newer systems such as digital video recorders deliver outstanding performance and flexibility that maximizes the value for the end user. Purpose-built to handle vast quantities of data, high-capacity digital video recorders allow security professionals to record and manage video data without having to sacrifice the high-quality imagery captured by advanced HD cameras. In addition, the IP capabilities of digital video recorders provide end users with additional flexibility, allowing them to scale their systems while minimizing costs associated with storage, maintenance and repairs. Ultimately, these innovations ensure that end users realize the greatest value for the long-term.
Maroussis: Storage is critical for any businesses’ security and after-the-fact coverage, and it’s one of the most costly considerations to any integrator and their clients. Luckily, with edge recording, trickling and growing Video Surveillance-as-a-Service and the Cloud - like our Stratocast system with Microsoft Windows Azure - storage options have gotten so much broader. The new storage and capacities with read /write controller, virtualization all have increased in performance and will continue to decrease in price.
SIW: Where will we see camera innovation come from over the next several years?
Nilsson: There is a lot of technology on the way that will improve cameras further. H.265, for example, is a compression that definitely will happen, but we need to remember with the previous compression, H.264, that was ratified in 2003 we had our first product in 2008 and it really became important in 2009 or 2010. There was really a six or seven year lag from the standard being ratified until it had a huge impact in the security industry. There are many reasons for that and one is it takes time for our chipset technology and sensors and all those kinds of things to catch up. H.265 will not take a long time, but it will take a couple of years. However, in the meanwhile, there will always be companies that try to get an edge from a marketing perspective by claiming they are the first ones to have something. You can re-tag a stream that is H.264 to call it H.265 even if you don’t get any benefit. So, just like any other time in this industry when there is a new compression standard, we’re going to see some confusion over a couple years where some company’s trying to gain an edge by I’m not saying lying about their products, but probably not telling the whole truth about it. The same thing with higher-resolution such as 4K, which is tied into this as well. That will also come about, but if you look at the cost of a lens that is really 4K compatible, you’re looking at a totally different level of price per camera and I’m not sure every customer wants to carry that. But, it is an interesting technology play, you get some excitement and some companies use that to their advantage.