Image Compression

 

1.      What is the code sequence for the following two zig-zag ordered 8x8 block (Assume that the first component in the previous block is 35; You will need the tables I used in the slides for this, go ahead and refer at them)
40, -8, 0, -4, 60 zeros;

2.      Answer

a.      Is it possible to take an image coded in JPEG format and extract just the DC components without reading (and decoding) every single byte in the file?

b.      The baseline sequential mode JPEG coding is a lossy coding mechanism. Which step of the coding process introduces the loss?

3.      If it takes 8 calculations to compute a 1-D DCT on an 8-element vector how many calculations does it take to do a 2-D DCT on an 8x8 matrix?

4.      Answer true or false

a.      When we compute the DCT of an 8x8 block of pixels the resulting F(0,0) component is called the AC component.

b.      In DCT, if we do not drop any frequency components, the inverse transform will give us the original function with no loss.

c.      Humans are more immune to loss of low frequency components.

d.      In the successive approximation technique we send bits of a pixel starting from LSB and working our way gradually towards the MSB.

e.      Lossless JPEG coding mechanism is based on DCT

f.       The JPEG Hierarchical mode can be used on the web to allow browsers to choose the most suitable resolution based on the browser capabilities.