This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: # 24 ECE 253a Digital Image Processing Pamela Cosman 12/9/11 ECE 253a Final Exam Solutions 1. Binary Image Processing: (8 points) (a) Even though the origin of B 1 is not centered, the set ( A ⊖ B 1 ) will be an outline of a square with side length 4 L with a single point in the center. (b) If B 2 has side length 2 L , then the result will have a solid interior. Grading: 4 points for part (a). Partial credit if you realize that the thick square erodes to a thin square outline, or that the inner circle erodes to a single point. 4 points for part (b). 2. Histogram Equalization: (3 points) First one is blue, second one is red, third one is green. Grading: one point each. 3. Huffman coding after DPCM: (10 points) (a) The difference values are − 3 and +3 (each has probability 1/16), − 2 and +2 (each has probability 2/16), − 1 and +1 (each has probability 3/16), and 0 (with probability 4/16). (b) The symbols are listed on the left, the probabilities next (multiplied by 16), and the codewords on the right. There are several different valid sets of Huffman codewords for this distribution. 4 3 3 2 2 1 1 +113 +32 +2 1111 1110 110 101 100 01 00 (c) The original symbols all have equal probability, so the average length of a Huffman code designed for them is 2 bits/symbol. The average length for the codewords found in part (b) is 1 2 × 4( 1 16 ) + 2 × 3( 2 16 ) + 3( 3 16 ) + 2( 3 16 ) + 2( 4 16 ) = 43 16 = 2 . 6875 bits/symbol The penalty is therefore 0.6875 bits/symbol. (d) Huffman coding the difference values is worse because the original values are independent. Grading: (a) 2 points, (b) 3 points, (c) 3 points, (d) 2 points. 4. Arithmetic encoding: (6 points) The encoder will put out a bit of zero after N consecutive zeros have been produced by the source, where N is the smallest integer such that (0 . 95) N < . 5 Grading: One point for saying that the bit will be zero, and 5 points for the inequality. Note: The initial assignment of source symbols to portions of the unit interval is arbitrary....
View
Full
Document
This note was uploaded on 01/15/2012 for the course ECE ece253A taught by Professor Cosman during the Fall '11 term at UCSD.
 Fall '11
 Cosman
 Image processing

Click to edit the document details