
Prel i m nary -14- November 1997
W9966CF/TF
Video Camera Interface Controller with Compression
7.2Video Encoding Engine (VEE)
The Video Encoding Engine is used to compress or bypass the CCD data, which is then stored in the
frame buffer for frame rate adjustment. CCD data may be bypassed or quantized, or compressed in
this block, depending on the input/output format, resolution, video quality, and parallel port bandwidth.
Two kinds of transfer mode are provided:
Original Transfer Mode :
CCD data is bypassed or quantized. Normally it is used when input data is
CCD raw data (bypass only), or when input data is in YUV 4:2:2 format with low resolution while very
high video quality is required. When input data is in YUV 4:2:2 format, the output data stream may be
in either YUV4:2:2 or YUV4:2:0 format. The output data stream is in YUV4:2:2 format when the
CR13_2 bit is 0, while it is in YUV4:2:0 format when the CR13_2 bit is 1.
Compressed Transfer Mode :
CCD data is compressed using proprietary algorithms. There are two
types of pictures used for compression ratio/SNR scalability: base layer and enhancement layer
pictures. Most of the compression algorithm is done for the base layer picture, which introduces
artifacts and distortions. The coding error picture can be encoded and sent to the host, producing an
enhancement layer picture to the decoded base layer picture. The extra data of the enhancement
layer picture serves to increase the signal-to-noise ratio (SNR) and also decrease the compression
ratio of the input video picture, and hence decrease the term compression ratio/SNR scalability. The
output data stream can be in YUV4:2:2 format only, the CR13_2 bit must be programmed 0.
Figure 7.2 shows functional block diagram for the VEE.
Base Layer
Line Buffer
CCD
Data
Base Layer
Encoder
Base Layer
Stream Buffer
Base Layer
Reconstruction
Enhancement Layer
Line Buffer
Enhancement Layer
Encoder
SRAM
B-FIFO
E-FIFO
SRAM
FIGURE 7.2 VIDEO ENCODING ENGINE (VEE) BLOCK DIAGRAM