Security and surveillance cameras continue to move from analog-based systems to digital, which lets companies move video over their existing IP-based data networks.

Sometimes, however, the images from these security cameras (both digital and analog) - especially in unusually bright or dark situations - often appear blurry, or are so saturated with light you can't see a thing. A security camera that provides a horrible image isn't really protecting your company, is it?

I recently met with officials from a company in Cambridge, Mass., that wants to change that. SMaL Camera Technologies (the S, M and L are the initials of the company's founders) recently announced a network camera kit (the V1200C Professional NetCam) that lets OEMs create network cameras that have intelligence built into them. These "smart cameras" will include SMaL's Autobrite technology, which improves the images from a security camera so that bright areas are not saturated and dark areas are not too dark. The Autobrite technology also is available from SMaL for analog cameras, company officials say.

Having a better picture lets companies get more from their security cameras. When you add motion-sensing or facial-recognition applications, having a picture that isn't washed out by light means those applications work even better.

In addition to the Autobrite technology, SMaL is offering network camera manufacturers image-recognition algorithms that let security applications (including facial recognition or motion sensing) integrate with a network camera. The idea, SMaL officials say, is to let the network camera do more of the work on the processing side, so that network bandwidth is not overutilized. In many network-based cameras, a lot of the processing is handled by an application on the computer, which means more network traffic between the computer and the camera. SMaL's technology allows this processing to be handled by the camera (with integration from the application).

The SMaL technology also provides a built-in server for User Datagram Protocol (UDP), HTTP and TCP/IP support, and can provide JPEG compression for images at rates up to 30 frame/sec with VGA resolution.

This marks SMaL's entry into its second market. Last year the company introduced the Ultra-Pocket VGA Digital Camera, which several companies have branded under their own names (including Oregon Scientific, Fuji and Logitech). These Ultra-Pocket digital cameras have the distinction of being the thinnest digital cameras available, and provide VGA-resolution images. SMaL says it's working on a 1.3-megapixel version of the Ultra-Pocket for distribution to network camera manufacturers later this year. SMaL also is working with automotive manufacturers to develop video technology in driver-assistance applications - imagine a car with a video camera on the rear bumper that shows what is directly behind the driver (via pop-up screen on or near the dashboard), and you'll get a feeling for what SMaL is working on. Another idea is to have cars that can sense whether a driver is drifting off the highway and then be able to shake the driver's seat as an alert. SMaL is working on the video camera technology for the application.