Question 10 (30%) a a. Sketch the block diagram of a standard Joint Picture Expert Group (JPEG) encoder, ignoring the initial block which subdivides an image into macroblocks (MB). Hence, explain in your own words the function of each of the three main JPEG blocks and where appropriate, include numerical values i.e., MB sizes. Clearly identify whether the processing in each block is either lossless or lossy. (10 marks) b. Figure 2 displays the Discrete Cosine Transform (DCT) coefficients for an 8x8 pixel MB in a standard JPEG encoder, where each DCT coefficient has been both quantised and had thresholding applied. 0 175 -1 -1 1 1 0 0 0 0 0 0 0 2 1 0 -1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Figure 2: Quantised DCT 8x8 pixel MB i. Determine the zig-zag pattern JPEG bitstream for this MB. ii. What is the dc coefficient value of this MB? iii. Explain the reason why the dc coefficient is encoded differently from the ac coefficients in the JPEG bitstream. iv. Given the level of DCT quantising/thresholding for the ac coefficients in this MB, estimate the compression ratio achieved. v. Run Length Encoding (RLE) is now applied to the ac coefficients in the JPEG bitstream obtained in Part (i). Assuming an End-of-Block code 0 0 is used, what will be the resulting RLE sequence? (12 marks) Motion-JPEG (M-JPEG) is a compression format which applies JPEG image coding independently to each frame of a video sequence. State why, given its relative simplicity, M-JPEG has not been more widely adopted as a video compression standard? (2 marks) Hence, with the aid of a suitable sketch, describe how motion vectors are used in all MPEG video coding standards to overcome the key limitation of M-JPEG. (6 marks) Question 10 (30%) a a. Sketch the block diagram of a standard Joint Picture Expert Group (JPEG) encoder, ignoring the initial block which subdivides an image into macroblocks (MB). Hence, explain in your own words the function of each of the three main JPEG blocks and where appropriate, include numerical values i.e., MB sizes. Clearly identify whether the processing in each block is either lossless or lossy. (10 marks) b. Figure 2 displays the Discrete Cosine Transform (DCT) coefficients for an 8x8 pixel MB in a standard JPEG encoder, where each DCT coefficient has been both quantised and had thresholding applied. 0 175 -1 -1 1 1 0 0 0 0 0 0 0 2 1 0 -1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Figure 2: Quantised DCT 8x8 pixel MB i. Determine the zig-zag pattern JPEG bitstream for this MB. ii. What is the dc coefficient value of this MB? iii. Explain the reason why the dc coefficient is encoded differently from the ac coefficients in the JPEG bitstream. iv. Given the level of DCT quantising/thresholding for the ac coefficients in this MB, estimate the compression ratio achieved. v. Run Length Encoding (RLE) is now applied to the ac coefficients in the JPEG bitstream obtained in Part (i). Assuming an End-of-Block code 0 0 is used, what will be the resulting RLE sequence? (12 marks) Motion-JPEG (M-JPEG) is a compression format which applies JPEG image coding independently to each frame of a video sequence. State why, given its relative simplicity, M-JPEG has not been more widely adopted as a video compression standard? (2 marks) Hence, with the aid of a suitable sketch, describe how motion vectors are used in all MPEG video coding standards to overcome the key limitation of M-JPEG. (6 marks)
