- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic to the Top
- Bookmark
- Subscribe
- Printer Friendly Page
2D DCT for image computatio n
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content
04-25-2012 11:30 AM
How to build 2D DCT for image computing
i am going to make watermarking model in system generator using DCT for that i need to build 2D DCT for 64*64 gray scale image.. and then going to apply watermarking algoritm to embbed binary secret image..
need help if somebody can provide some literature or any material.
Thnks
Re: 2D DCT for image computatio n
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content
04-25-2012 01:23 PM
Hi.
So you need to implement 64x64 2D DCT.
First of all, design input/output interfaces for your model. Perhaps the following thread will be helpful:
Concerning 2D DCT, it can be calculated applying 1D DCT to all rows and then to all columns (or vice versa) of input image.
There are different ways of implementing 1D DCT. 1D DCT is essentially array of correlations of input signal and cosine functions.
For example, you can use multipliers and directly calculate sums of products. Or you can use 1D FFT to calculate 1D DCT (by selecting only real outputs of FFT).
Vitaly.
Re: 2D DCT for image computatio n
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content
05-15-2012 03:57 PM
thnks vitaly...
vlavruhin wrote:
Hi.
So you need to implement 64x64 2D DCT.
First of all, design input/output interfaces for your model. Perhaps the following thread will be helpful:
Concerning 2D DCT, it can be calculated applying 1D DCT to all rows and then to all columns (or vice versa) of input image.
There are different ways of implementing 1D DCT. 1D DCT is essentially array of correlations of input signal and cosine functions.
For example, you can use multipliers and directly calculate sums of products. Or you can use 1D FFT to calculate 1D DCT (by selecting only real outputs of FFT).











