KSC-3000 Kodak, KSC-3000 Datasheet - Page 4

• Manufacturer Part Number: KSC-3000
• Description: Color Processor Firmware Code Providing 24-bit RGB Color Image Data at 1 Megapixel Resolution And 30 Frames/second.
• Manufacturer: Kodak

IMAGE SENSOR SOLUTIONS
Figure 2 - Xilinx Virtex II Overview
Functional Description
The KSC-3000 is image-processing code designed to work with the Kodak KAI-1020CM interline CCD, which
incorporate color filter arrays. The KSC-3000 provides all the processing steps required to convert the raw image
into a full color image at 30 frames per second. The KSC-3000 provides a number of controls that allow the user to
set various input image parameters to control the processing done on the image.
The KSC-3000 has three interfaces. The first is a three wire serial interface. This interface allows setup, algorithm
control and configuration, and status read back. The second is the input data interface. The interface consists of the
10 bit data values, a frame valid indication, and a line valid indication. The third is the output data interface. The
interface consists of 24 bit data values (8 bits each for red, green, and blue), a frame valid indication, and a line valid
indication.
The KSC-3000 datapath is illustrated in Figure 3, a brief description of each block is given here. Most digital
imagers use a single color image sensor that provides only one color value for each pixel. In this case, the pixels on
the image sensor are covered with a mosaic of transmissive filters (typically red, green and blue), commonly known
as a color filter array (CFA). Many digital cameras use the "Bayer" pattern CFA, named after its inventor, Kodak
scientist Bryce Bayer. In the "Bayer" CFA pattern, 50% of the pixels are green, 25% are red, and 25% are blue.
(The human eye gets most of its sharpness information from green light, which is why the CFA pattern was
designed to have more green pixels.)
To get a good color image, the processing must provide white balancing. While both daylight and indoor lighting
provide "white" light, daylight actually has more energy in the blue portion of the light spectrum and indoor lighting
provides more energy in the red portion of the spectrum. Your eyes and brain automatically adapt to this difference,
so that a white piece of paper appears white under all normal lighting conditions. An image sensor, by itself, cannot
compensate for the illumination. Therefore, the image-processing datapath allows the user to adjust the red and blue
signal strengths to match the green signal strength in white and neutral areas of the picture.
To create a complete full-color image, a digital image-processing step called CFA interpolation is used to "fill in"
the missing color values for each pixel. Interpolation takes the sequence of single-channel color pixels and creates a
full three-channel RGB color image. A sophisticated algorithm decides if the "missing" color values are in a smooth
K S C - 3 0 0 0 R e v 1 . 0
4
w w w . k o d a k . c o m / g o / i m a g e r s
5 8 5 - 7 2 2 - 4 3 8 5
E m a i l : i m a g e r s @ k o d a k . c o m