Advancements in intelligence at the edge have enhanced the value proposition of video surveillance cameras.
Photo credit: (Image courtesy bigstockphoto.com)
John Grabowski is national sales and marketing manager for JVC's Security Division.
Photo credit: (Photo courtesy JVC)
In the past, video surveillance cameras were typically perceived as one component of a larger security system made up of many parts and pieces that captured video which may or may not have been used later – if it were usable at all. Today, this perception no longer stands, as cameras now deliver greater functionality and better overall performance as a result of breakthroughs in imaging and processing technologies, and the emergence of on-board intelligence.
The two image sensor types available today, CCD and CMOS, are driven by incredibly robust processing power that allow the sensors to analyze images pixel-by-pixel to produce high-quality images with excellent resolution and color rendition. The most powerful advances in intelligent camera features have come in the following areas:
Wide Dynamic Range
Wide dynamic range (WDR) allows cameras to capture much more detail in scenes where there are varying levels of light, much like the way the human eye "processes" these types of scenes. With WDR, usable images can be captured when there are both very bright and very dark areas within the same scene. These conditions are common in building lobbies with large windows, as well as in showrooms and warehouse loading docks. Similarly, cameras placed in outdoor areas such as parking lots - where bright headlights can wash out an entire scene - now benefit from WDR.
Security never sleeps, so cameras have to be able to work around-the-clock, regardless of the amount of available light. True day/night functionality has come a long way in recent years, to the point where some cameras have the capability of automatically switching from color to black-and-white operation based on available light. Even more advanced technologies allow some cameras to deliver color images in low-light situations without the need for IR illuminators or other external lighting sources. Not only does this allow cameras to be deployed in areas with changing lighting conditions, it means that color images can be obtained for identification purposes 24/7.
Cameras with embedded video analytics, such as motion, face, and object detection, analyze image data at the point of capture and can effectively eliminate the need to transmit video and data to a central server. This enables very efficient use of both transmission and recording bandwidth. Using analytics, some cameras can also be set to record video at a lower resolution and/or frame rate, and then automatically increase resolution and frame rate to capture higher-quality video when triggered by an event, such as motion in a scene or particular area of interest. By placing intelligence on the edge, users can save substantially on expensive centralized analytics software and licensing fees that may only be required in specific areas and camera locations.
New system models are being developed that push recording functions to the edge as opposed to centralized servers. One way this is being achieved combines the use of new IP surveillance cameras with on-board SD card recording and analytics along with local NVRs and new video management system (VMS) solutions. This virtually eliminates the need for centralized servers by pushing live and/or recorded video triggered by an event such as motion detection, while maintaining a critical back-up of recorded events even in the event of a network failure.
Regions of Interest
Because extraneous motion, reflections and other conditions within a camera’s view can increase bandwidth and storage requirements, it’s not always ideal to record a camera’s entire field-of-view. Fortunately, many cameras allow users to select the highest-priority areas within a scene, such as doorways, and record them at the highest resolution. Other areas can be recorded at lower resolutions, reducing the amount of detail in the scene and thereby minimizing bandwidth and storage requirements and the costs associated with them.
Camera sensors can pick up noise or interference caused by a camera’s circuitry and electrical components, which then shows up in footage as static, snow or fuzzy images that increase the detail, and by extension, bandwidth requirements. By using sophisticated computing processes to average out the pixels in a frame, cameras are capable of reducing noise in captured video. Highly advanced 3D noise reduction (3DNR) technology provides an even more powerful way to capture clearer images in dark environments by comparing continuous frames and the noise data blended over time.
The primary culprit behind increased bandwidth and storage costs is resolution, which is also the main reason video surveillance cameras have become so appealing to end users. Fortunately, this issue has been somewhat mitigated by video compression technologies like H.264 and MPEG. Nearly all cameras include some form of compression capability, using a complex algorithm to compress video data into smaller file sizes, which require much less bandwidth for transmission.
These and other improvements in camera intelligence have dramatically changed the perception – and value proposition – of video surveillance cameras, transforming what may have once been considered a system component purchase into a valuable system solution that delivers a substantial ROI. As manufacturers continue to pack even more intelligence into cameras, what comes next is anyone’s guess. One thing is for sure: end users will reap the benefits of improved security and an even greater value proposition from highly intelligent cameras.