Home

Linescan cameras target security applications

Many CCTV video cameras provide a resolution of 640 × 480 pixels.
March 1, 2003
3 min read

Many CCTV video cameras provide a resolution of 640 × 480 pixels. To confidently identify people within these images, the resolution of these cameras must approach 40 pixels per foot to provide a maximum field of view of 16 × 12 feet. To save storage, digital video recorders are often used to reduce the image size to 320 × 240 pixels and decrease the field of view to 8 ft wide.

To overcome the limitations of these CCTV products, Spectrum San Diego (San Diego, CA; www.sentryscope.com) has used linescan technology to increase camera resolution. Linescan camera technology has been in use for many years in web-inspection applications, using a single line of pixels to build up images of moving objects.

To adapt linescan technology to surveillance applications, the vertical scan line of a CCD camera must be swept across the field of view to build up an image. The SentryScope from Spectrum uses a 2048 vertical-pixel Eclipse CCD cameras from Dalsa (Waterloo, Ontario, Canada; www.dalsa.com) and an optical assembly that uses a motor from Emoteq Computer Optical Products (Chatsworth, CA) to rotate a front surface mirror across the field of view. With this approach, the camera system collects up to 10,240 vertical scan lines for a total of nearly 21 million pixels in a single image.

Capturing up to 10,240 pixels in the horizontal direction yields a field of view of up to 250 ft wide with a single camera. But because of the high number of pixels captured in the linescan implementation, a typical rate of one image per second is achieved by the SentryScope.

To control the operation of the mirror and the camera, Spectrum uses a separate digital controller board based on the 600-MOPS BlackFin DSP from Analog Devices (Norwood, MA; www.analog.com). The scanning assembly, custom designed by Spectrum, is digitally controlled to provide the required accuracy and precision. After images are buffered into a 64-Mbyte SRAM on the DSP controller board, the camera and the PC transmit the compressed data stream over a 100-Mbits/s Ethernet link to a 2-GHz PC with a storage capacity of up to 480 Gbytes.

The target application for SentryScope is surveying a large indoor or outdoor area that cannot be monitored 100% of the time with existing CCTV cameras and where personnel are not available to take immediate action in case of an incident (see figure). At one image per second in black and white, monitoring live images probably will not be the most common use of the camera system.

When reviewing stored images, images captured from high-resolution cameras such as the SentryScope can clearly identify faces and read license plates. Other possible camera applications include the large-area inspection of PCBs where large images must be captured at high resolution.
Click here to enlarge image

null

Partially manned and unmanned operations, where the review of stored images is a regular occurrence, are expected. Other uses for the camera could be in machine vision, where large areas of PCBs may need to be scanned at high-resolution.

Sign up for Vision Systems Design Newsletters

Voice Your Opinion!

To join the conversation, and become an exclusive member of Vision Systems Design, create an account today!