readme.txt

code for video compression using simulink

Simulink-based video processing demos using DSP Blockset.
All demos are based on R13 (MATLAB 6.5)

1: vid_dct.mdl
   DCT-based image compression
   Computes 2D-DCT of each video frame, then removes all DCT coefficients
   that have magnitude below a threshold.  This is the basis of image compression.
   The original and reconstructed video frames are shown as the video runs.
   
2: vid_focus.mdl
   Shows real-time computation of 2-D FFTs on video frames.  The video contains
   a microscope image coming in and out of focus.  A comparison of the high-spatial
   frequency content to the low spatial-frequency content of each video frame
   gives an indication of the relative focus adjustment; a plot of this ratio as
   the video comes into and out of focus shows how well this metric works.
   
   In practice, an FFT would not be needed (two 2-D filters would be employed).
   However, it is educational to see the 2-D FFT results to gain an appreciation
   of the effect of focus on high spatial-frequency content of video frames.