Ordering number : EN*5402
CMOS LSI
LC82103
Image-Processing LSI for Facsimile, Copier,
and OCR Applications
Preliminaly
Overview
The LC82103 converts the analog video signal from the
CCD contact image sensor into high-precision binary
image data. The LC82103 includes an 8-bit A/D converter
and two 6-bit D/A converters. The 6-bit D/A converters
are used for setting reference potential (high and low) of
the 8-bit A/D converter. The LC82103 converts the input
analog data to high precision multilevel data, that is
produced by all pixel shading correction process, multi-
level image resolution converting process, and
緯
correction process of user defined curves. That multilevel
data is converted to high-precision bi-level image that is
produced by using two dimensional filtering, a image
separation method, and an error diffusion technique.
Finally the LC82103 is able to reduce the image in the
horizontal and vertical directions, too. The LC82103
requires absolutely no external memory since it limits the
number of pixels processed to 3072/line. This LSI can
implement the image processing used by FAX, copier, and
OCR applications.
鈥?Binary image reduction (horizontal: thinning, fine black
line retaining, fine white line retaining, vertical:
thinning, fine black line retaining)
鈥?Single-voltage 5 V supply and low power due to CMOS
process fabrication
Package Dimensions
unit: mm
3159-QFP64E
[LC82103]
Features
鈥?Number of pixels processed: 3072 pixels/line
鈥?Processing speed: 250 ns/pixel maximum (When
CLKIN = 32 MHz)
鈥?Built-in 8-bit A/D converter (with a sensor signal timing
adjustment function)
鈥?Built-in two 6-bit D/A converters for setting the A/D
converter high and low reference potential
鈥?Sensor drive circuit (supports CCDs and several CIS
types)
鈥?Digital clamping (single-point clamping, even/odd
clamping)
鈥?Distortion correction (white correction: all pixels, black
correction: settable black correction subtracted data)
鈥?/div>
緯
correction (support user-defined curves: 8-bit data)
鈥?Image area separation (characters, photographs,
screened halftone areas)
鈥?Simple binary coding process (fixed threshold,
brightness adaptive threshold)
鈥?Intermediate processing error diffusion (64 levels)
鈥?Multilevel image resolution conversion (1/2, 2/3, 3/2,
2/1)
SANYO: QFP64E
SANYO Electric Co.,Ltd. Semiconductor Bussiness Headquarters
TOKYO OFFICE Tokyo Bldg., 1-10, 1 Chome, Ueno, Taito-ku, TOKYO, 110 JAPAN
63096HA (OT) No. 5402-1/5
1
2
3
4
5
