AnsweredAssumed Answered

ADV212 compression rate

Question asked by chengyulan2008@sina.com on Aug 25, 2015
Latest reply on Aug 26, 2015 by DaveD


By using the single-chip  JPEC2000  codes (ADV212), a real-time compression system is presented for high resolution image.

 

ADV212 Firmware Parameter Encode Settings

 

Name

Parameter Description

Bytes

Settings

1

VFORMAT

Video Standard

1

Custom-specific :0X04

2

PREC

Precision

1

12bits :0X02

3

XFORMLEV

Number of Wavelet Transform

1

5 levels of transform :0X05

4

UNI

Component Polarity

1

Unipolar C, Unipolar Y :0X03

5

CBSIZE

Code Block Dimensions

1

128*32:0X03

6

WKERNEL

Wavelet Kernel

1

Irreversible 9*7:0X00

7

STALLPAR

Skip Fields in Encode Mode

1

Capture all fields:0X00

8

ATTRTYPE

Attribute Data Output Format

1

Do not output attribute data:0X00

9

RCTYPE

Rate Control Algorithm

1

Within 5% of RCVAL:0X01

10

RCVAL

Target Size or Quality Factor

3

XTOT * YTOT * PRES / Rate / 8

11

J2KPROG

JPEG2000 Progression Style

1

LRCP :0X00

12

PICFG

Pixel Interface

1

0X0F

13

QFACT

Quantization Factor

1

256/256:0X00

14

COD_STYLE

Output Code Stream Format

1

0X01

 

The compression rate is 4, XTOT is 2048,YTOT is 180,so RCVAL= XTOT * YTOT * PRES / Rate / 8=2048*180*12/4/8.

Issue:

If the input image source is random-data,the result of compression rate is 4,but if the input image source is correlative image source,the result of compression rate far greater than 4.Why?

Why the result of compression rate is variational,not fixed(RCVAL= XTOT * YTOT * PRES / Rate / 8=2048*180*12/4/8)?

Outcomes