User manual

Software Platform
XAPP794 (v1.3) December 20, 2013 www.xilinx.com 42
Figure 46 illustrates how the image statistics handler interacts with the image processing
pipeline.
The image processing pipeline consists of these IP cores:
Defective pixel correction
Color filter array interpolation
Image statistics
Color correction
Gamma correction
Image enhancement
The image statistics IP core provides an important link with the Linux applications. It gathers
statistics about the camera input, including intensity and chroma histograms, which are used to
automatically adjust the image processing pipeline to improve image quality.
The image statistics task provides some interesting application examples that are all based on
the image statistics IP core:
Auto white balance: Attempts to determine the lighting conditions and applies a weighted
sum of one or more of the pre-determined white balance corrections based on the
calculated probability of each of the possible light sources using the color correction matrix
IP core.
Auto-gain: Based on the intensity of the image, reconfigures the image sensor gain
setting.
Auto-exposure: Based on the intensity of the image, reconfigures the image sensor
exposure setting.
Global contrast: Automatically determines a gamma correction look-up table.
X-Ref Target - Figure 46
Figure 46: Image Statistics Handler
*DLQ
&RQWURO
([SRVXUH
&RQWURO
&DPHUD
,QSXW
3URFHVVLQJ6\VWHP
3URJUDPPDEOH/RJLF
$;,/LWH
;BB
$XWR
*DLQ
$XWR
([SRVXUH
,PDJH6WDWLVWLFV+DQGOHU
,PDJH
6WDWLVWLFV
+DQGOHU
$XWR
:KLWH
%DODQFH
*OREDO
&RQWUDVW
'HIHFWLYH
3L[HO
&RUUHFWLRQ
&RORU)LOWHU
$UUD\
,QWHUSRODWLRQ
,PDJH
6WDWLVWLFV
&RORU
&RUUHFWLRQ
*DPPD
&RUUHFWLRQ
,PDJH
(QKDQFHPHQW