RGB-only sensors’ future threatened by RGB+IR Fusion which offers highly accurate color reproduction and high resolution
UK image processing software specialist Spectral Edge is launching its RGB (visible light) and Near Infrared (NIR) Fusion solution ‘Spectral Edge Fusion’ into the surveillance market at ISC West 2019 to be held in Sands Expo, Las Vegas this week. An RGB-only sensor will be live demonstrated alongside an RGB+IR sensor with Spectral Edge Fusion on board, on Booth Number 41030 in the Emerging Technology Zone at the show.
Spectral Edge Fusion can provide surveillance cameras with high quality, color-accurate images even in low or mixed lighting conditions and shadows. It is also able to capture much more depth and color in detail that would normally be hidden in foggy or hazy conditions.
Crucially, as well as delivering better-looking images to human viewers, the Spectral Edge Fusion platform enables applications such as facial or object recognition to deliver much higher success rates in low light and shadowy conditions.
Spectral Edge Fusion is particularly useful in supporting video analytics capabilities such as facial, object and action recognition, as well as other surveillance applications which demand high-quality images to achieve accurate identification to minimize false positive alerts.
Spectral Edge Fusion uses its patented fusion algorithm to create three channels from four original channels: Red, Green, Blue and NIR light. It’s the only technology on the market today which can perform this light fusion without introducing pixelation, artifacts or additional noise. Spectral Edge predicts its development in RGB and NIR fusion will render RGB-only sensors obsolete within the next few years once the technology is proven in the field.
Spectral Edge Fusion also ensures minimal loss of resolution – while delivering improved contrast, dynamic range and signal-to-noise ratio. So, what’s lost in color as daylight fades is made up for in additional texture and depth of image which the NIR delivers given suitable IR illumination.
It also generates a Bill of Materials saving for surveillance camera manufacturers because only one readily-available sensor and lens is needed for a single day/night camera. No mechanical switching between sensors is required and there is no requirement for an IR-cut filter. This helps to minimize moving parts which wear out over time and extends surveillance cameras’ Mean Time to Failure.
An additional cost advantage flows from using NIR because it can be detected by the same sensor chip as RGB. There is no extra cost associated with such manufacturing variants. Put simply, some of the pixels which sense visible light are replaced with pixels sensitive to NIR. Spectral Edge Fusion has already been successfully tested on a range of such sensors which support both RGB and NIR wavelengths. Spectral Edge Fusion can be fitted into System-on-Chips (SoCs) which process sensor outputs without increasing the silicon area significantly.
The innovator will show RGB-only sensor video alongside video produced by an RGB-IR sensor with Spectral Edge Fusion technology on board, both supported by the same level of IR illumination, to show chip and camera manufacturers the much better image quality which is possible using RGB-IR sensors and Spectral Edge Fusion technology in lower lighting conditions, at its booth (number 41030) at the show.
Rhodri Thomas, Chief Executive Officer, Spectral Edge, said:
“We think that our innovation in RGB+NIR fusion could render RGB-only sensors obsolete a few years from now. We are looking forward to talking about this with surveillance market players in the US where so many great camera and chip manufacturers operate. The market here is highly-focused on manufacturing high-quality sensors designed to optimize the latest video analytics innovations. It is these OEMs, system builders and integrators which Spectral Edge hopes to meet at ISC West this week.”
Spectral Edge has just published its latest White Paper which goes into more detail about how Spectral Edge Fusion works. To read more follow this link.