A/D Converter: A device that accepts an analog input signal and converts it to a digital signal. The process requires that the analog signal be broken into a discrete number of digital values, each representing a range of analog values. The number of digital values is limited, usually 256. One result of this quantization is that analog values between two adjacent digital values are lost. For this reason, the Light Science Technologies three-dimensional image enhancement system works best with an analog signal input.

Archive: Permanent storage hard disk. This can be on the Light Science Technologies three-dimensional image enhancement system workstation, or a laboratory server.

Background: The portion of an image where no subject is present.

Contrast: The difference between the brightness of subject’s feature edges and the average brightness of the background.

De-enhance: Blend into the background.

Depth of Field: The portion of the subject’s physical depth that is in focus. Generally, the depth of field of one image is substantially less than the subject’s physical depth, leaving a good portion of the subject out of focus.

Edges: In a two-dimensional image, edges are where the subject stops and the background starts.

Emphasize: Contrast from the background

Enhance: Contrast portions of a feature of interest.

Features of interest: Portion of a subject that is interesting.

Features of no interest: Portions of a subject or background that are of no interest.

Floating Point Calculations: Calculations performed on floating point numbers rather than integers. Floating-point calculations take many more machine cycles than numeric calculations.

Focal Plane: A two-dimensional image plane of a three-dimensional subject. In any given focal plane only the portion of the subject contained within the depth of field will be in focus. The remainder of the subject will be out of focus.

Image Capture: Process of digitizing an analog image and storing it digitally in memory.

Image Frame: A single captured image.

Imaging Tool: Laboratory instruments that produce images as their output product. These include microscopes, telescopes, x-ray equipment, ultrasonic imaging equipment, etc.

Interframe Relationships: Analysis of pixel values, image to image, that are at the same x-y location on each image plane.

Integrate: The process of averaging pixel values over image planes that are at the same x-y location on each of the image planes.

Laboratory Server: Computer providing shared resources in a laboratory over a LAN.

LAN: Local Area Network.

Noise: The background signal level at which no more useful information can be extracted.

Numeric Calculations: Calculations performed on integers rather than floating-point numbers.

Pipeline Video Processing: Pipeline video processing is where images are processed sequentially by one processor, then passed on to the next, then passed on to the next something like a "bucket brigade." Pipeline video processing can only be effectively done with a supercomputer with multiple parallel processors. The Video DSP chip is particularly well suited for this processing because each of the parallel processors utilize a common memory. As a result, passing an image from one processor to another does not involve physically moving the image in memory, rather only signaling from one processor to the next.

Real Time: The Image Capture rate that is visually contiguous.

Resolution: Ability to resolve or see edges of a feature.

Signal: The video information produced by the subject.

Signal to Noise Ratio: The ratio of the noise video level to the signal video level.

Shades of Gray: Number of quantization steps used in the A/D conversion process, typically 256.

Subject: Portion of the image that constitutes the reason the image was taken in the first place.

Supercomputer: Generally, a computer is considered a supercomputer when it has an instruction speed in excess of 1 billion operations per second. Most supercomputers today use multiprocessors to attain these computational speeds. The KRELL™ series of supercomputers not only use multiple processors, it also utilizes a video pipeline architecture making it more suitable for video processing than conventional supercomputers.