Applications of digital signal processing - Continuous and Discrete Time Signals and Systems - page 784

For spatial location m = 0, n = 3, the predicted sample value, the prediction
error, the quantized error, and the reconstructed sample value are given by

i [0 , 3] = i

′

predicted value [0 , 2] = 153;

e [0 , 3] = i [0 , 3] − i [0 , 3] = 149 − 153 =− 4; error

e [0 , 3] = round( − 4 / 3) =− 1;

quantized error

[0 , 3] = i [0 , 3] + 3 e [0 , 3] = 153 − 3 = 150 .

′

reconstructed value i

For spatial location m = 1, n = 0, the predicted sample value, the prediction
error, the quantized error, and the reconstructed sample value are given by

i [1 , 0] = i

′

predicted value [0 , 0] = 156;

e [1 , 0] = i [1 , 0] − i [1 , 0] = 156 − 156 = 0; error

quantized error e [1 , 0] = round(0 / 3) = 0;

′

[1 , 0] = i [1 , 0] + 3 e [1 , 0] = 156 + 0 = 156 .

reconstructed value i

For spatial location m = 1, n = 1, the predicted sample value, the prediction
error, the quantized error, and the reconstructed sample value are given by

i [1 , 1] = 0 . 33( i

′ ′ ′

[1 , 0] + i [0 , 1] + i [0 , 0]) predicted value

= 0 . 33 468 = 154 . 44;

e [1 , 1] = i [1 , 1] − i [1 , 1] = 159 − 154 . 44 = 4 . 56; error

quantized error e [1 , 1] = round(4 . 56 / 3) = 2;

[1 , 1] = i [1 , 1] + 3 e [1 , 1] = 154 . 44 + 6 = 160 . 44 .

′

reconstructed value i

For spatial location m = 1, n = 2, the predicted sample value, the prediction
error, the quantized error, and the reconstructed sample value are given by

i [1 , 2] = 0 . 33( i

′ ′ ′

predicted value [1 , 1] + i [0 , 2] + i [0 , 1])

= 0 . 33 469 . 44 = 154 . 92;

e [1 , 2] = i [1 , 2] − i [1 , 2] = 159 − 154 . 92 = 4 . 08; error

quantized error e [1 , 2] = round(4 . 08 / 3) = 1;

[1 , 2] = i [1 , 2] + 3 e [1 , 2] = 154 . 92 + 3 = 157 . 92 .

′

reconstructed value i

For spatial location m = 1, n = 3, the predicted sample value, the prediction
error, the quantized error, and the reconstructed sample value are given by

i [1 , 3] = 0 . 33( i

′ ′ ′

predicted value [1 , 2] + i [0 , 3] + i [0 , 2])

= 0 . 33 460 . 92 = 152 . 10;

e [1 , 3] = i [1 , 3] − i [1 , 3] = 155 − 152 . 10 = 2 . 90; error

quantized error e [1 , 3] = round(2 . 90 / 3) = 1;

[1 , 3] = i [1 , 3] + 3 e [1 , 3] = 152 . 10 + 3 = 155 . 10 .

′

reconstructed value i