Tech Trends: A Panoramic View

Aug. 26, 2014
Advances in image processing has breathed life into the panoramic IP camera market

It is no secret that megapixel cameras have taken the surveillance market by storm — providing the ability to see more with fewer cameras, with progressively better video quality and digital PTZ features. Conversely, fish eye lenses providing up to 360 degree images of varying quality have been around for years. It is not surprising that the combination of high pixel density sensors and appropriate lenses, enhanced by sophisticated image processing algorithms, has led to expanded and enhanced product offerings in 180 degree and 360 degree panoramic IP cameras.

Let’s take a closer look at these three elements to see how they have combined to create this growing class of products.

 Sensors and Sensor Count

More image pixels provide more detail of the scene, but how big a sensor is needed and how many sensors are used. Two pixels per target may be sufficient for detection, while 24 pixels or more per target may be needed for recognition with high accuracy.

Axis Communications’ panoramic line uses a single sensor starting with a 1/3-inch progressive scan CCD 1.3 megapixel sensor coupled with unique lens technology and moving to a 5 MP progressive scan CMOS sensor (CMOS sensors provide the opportunity for individual pixel management). Samsung employs a 3 MP CMOS sensor while Mobotix uses 5 MP CMOS.

All rely on the lens system to deliver the full scene of images to a single imager, likely leading to lower average camera price points. Upstart Scallop Imaging uses 5 x 1.3 MP CMOS sensors, each with its own lens. While most all panoramic cameras are domes or mini-domes, Scallop staggers its lenses in a fashion that leaves the camera body relatively thin and flat, creating an unobtrusive package.

For increasing pixels on the scene, look at the Avigilon line, which employs 4 x 2 MP CMOS sensors, individually lensed, and spaced at 45- or 90-degree intervals around the circumference of the dome. Arecont Vision spaces its sensors the same way with individual image sensor options of 2 MP to 10 MP. Check out the side-by-side comparison in the chart below:

Don’t be surprised to see these numbers increase as the megapixel war continues and complementary technologies get better.

Lenses

Every image sensor has an associated lens. Where there is one sensor, the selection of lens will make or break the overall image quality. Immervision is one company who has specialized in lenses for panoramic cameras. In its literature, the company notes that “distortion is a design parameter.” By working with camera vendors and understanding the combined optical system, they create a panoramic video viewing library to allow the received video to be viewed in an optimized fashion.
Another approach is a two-lens system, as illustrated in the Axis P5544 PTZ dome. When in 360-degree mode, the camera functions as a normal PTZ dome with an 18x optical zoom; however, when looking straight down, a 2.7 mm fixed “Panopsis” lens precedes the regular zoom lens for the 360-degree view.

 Algorithms

Once you get past the optical system, it is time for the software magic. This occurs in three general areas: image correction, image display and image analysis.

Image correction can take several forms. If the device has a single imager and lens, dewarping and distortion management, such as with Immervision’s offering, optimizes viewability. If, on the other hand, multiple sensor/lens combinations are used, the individual image sections require stitching to provide a unified picture. Ideally, the process should present a seamless image with no obvious artifacts or discontinuities. Beauty is in the eyes of the beholder, so I would ask for a live demo to make a decision.

Multiple megapixels create lots of image data, which presents opportunities for multiple viewing windows — each with digital PTZ and analytics opportunities. It is common in many of these camera systems to have a rectangular wide screen view displayed alongside the windows of interest. Functionality comes down to the number of pixels available, effectiveness of the software and built-in processing power of the camera.

 Bandwidth and Aesthetics

Most of the better cameras offer various options for storage (local and remote) and transmission of the video data. These may include various frame rates and resolutions for selection areas of the scene, while viewing at different rates. Transmission, too, may be based on selected scenes or rules, transmitting only when or what is necessary. That’s important, because of the potential size of the data stream.

Finally, do not ignore aesthetics — particularly for a lower ceiling, these devices may be highly visible, so packaging and mounting options may be key selection criteria. All in all, this class of IP video cameras represents a nice addition to the array of product choices now available to the designer and integrator.

Ray Coulombe is Founder and Managing Director of SecuritySpecifiers.com and RepsForSecurity.com. He can be reached at [email protected], through LinkedIn at www.linkedin.com/in/raycoulombe or followed on Twitter @RayCoulombe.