ICM320T Datasheet, PDF (10/30 Pages) List of Unclassifed Manufacturers – 3.2 Megapixel QXGA Digital Color CMOS Image Sensor

ICM320T
3.2 Megapixel QXGA Digital Color CMOS VGA Image Sensor
Preliminary Data Sheet V1.0
Exposure Time
To accommodate different illumination
requirements, you can change the exposure time
on the ICM320T by adjusting registers 0x1C and
0x1D. The exposure time is measured in terms of
the time to read out one line of data.
The time for processing one line is equal to the
pixel clock period times the frame width. The
frame width is stored in registers 0x0C and
0x0D. The exposure time for one full resolution
default line width of 2200 pixel clocks is 34 µs
when the pixel clock is 64 MHz.
Digital Gain Control
The ICM320T digital gain control feature is one
of the methods you can use to control image
quality. Adjusting the digital gain is generally
used for minor changes, either to balance color
or to adjust the overall image luminance. The
gain range is from 1/256 to 8 of the four Bayer
pattern pixels.
Gain changes require changes to two types of
registers; the per Bayer color channel gain value
registers 0x20-0x28, and the magnitude of
change register 0x52[3:0].
Analog Gain Control
The ICM320T supports up to 4x analog gain
control for low light operation. Four analog gain
settings are programmable using register 0x28.
The default value 0x00, corresponds to 1x gain,
which is used for normal light. Other settings are
0x81: 2x gain, 0x92: 3x gain and 0xA3: 4x gain.
Modified analog gain settings are activated by
writing 0x1 into register 0x00[7].
Analog Gain Calibration
The analog gain can be calibrated by adjusting
the maximum analog input and the maximum
ADC output so that they have the same dynamic
range. To enable automatic gain calibration,
write 0x1 into register 0x00[1].
Pixel Timing and Function Control
The ICM320T has a software-controlled pixel
analog data path. This feature allows fine-tuning
of the sensor’s performance and makes the
sensor highly adaptable to a wide range of
applications. The pixel analog data path timing is
controlled by the default settings or by
downloading a pixel configuration table.
Register 0x02[4] makes the selection. Pixel
configuration tables are sequentially downloaded
into registers 0x04, 0x05, and 0x06 while
incrementing the 0x03 address pointer.
Subsampling Schemes
The ICM320T supports subsampling for
viewfinder and other reduced data rate output
modes. It supports sub-sampled resolutions in
the 1/4-rate 1024x768 and 1/16-rate 512x384
formats. The 1/4-rate mode maximum frame
rate is twice the maximum full resolution frame
rate. The peak 1/16-rate frame rate is four times
the maximum full resolution frame rate.
Subsampling modes are selected through register
0x52[7:4].
Dark Level Compensation
The ICM320T supports automatic dark level
compensation with programmable offset. The
automatic dark level compensation algorithm
supports user programmable hysteresis to
prevent oscillation of automatic exposure
algorithms in companion devices. The hysteresis
parameters are set using register 0x3B[2:0]. The
algorithm in enabled by register 0x3B[7].
Dead Pixel Filtering
The ICM320T supports dead pixel filtering for
both full QXGA resolution and sub-sampled 1/4-
rate and 1/16-rate resolutions. The filter is only
applied to pixels outside the high and low dead
pixel filter thresholds defined in registers 0x42
through 0x45. The filter is enabled and disabled
using register 0x51[3].
Pixel Noise Filtering
The ICM320T implements pixel noise
suppression filtering for improved image quality
under low light conditions for both full QXGA
resolution and sub-sampled 1/4-rate and 1/16-
rate resolutions. The pixel noise suppression
filter implements a two or four pixel averaging
across adjacent pixels of the same color space in
sub sampled resolutions. Register 0x51[2:0] is
used to enable and disable two pixel or four pixel
averaging respectively.
Copyright 2005, IC Media Corporation
http://www.ic-media.com/
Page 10
3/14/2005