Blind Measurement of Blocking Artifact in Images Zhou Wang Lab for - - PowerPoint PPT Presentation

▶

Nov 16, 2023 1.01k likes •1.21k views

Blind Measurement of Blocking Artifact in Images Zhou Wang Lab for Image and Video Engineering Dept. of Electrical and Computer Engineering The University of Texas at Austin 1 Blocking Effect Most image/video coding standards use DCT

SLIDE 1

Blind Measurement of Blocking Artifact in Images

Zhou Wang

Lab for Image and Video Engineering

Dept. of Electrical and Computer Engineering

The University of Texas at Austin

SLIDE 2

Blocking Effect

➘ Most image/video coding

standards use DCT

➘ Quantization is used to

achieve low bit rates

➘ Decoding is lossy ➘ Coding artifacts including

blocking, blurring, and ringing, etc.

➘ Blocking effect is usually

the most significant

SLIDE 3

Blocking Effect Measurement

✔ Applications

➘ Encoder - optimize parameter selection and bit allocation ➘ Decoder - design post-processing algorithm

✔Measurement Methods

➘ Raw numerical errors - Mean Squared Error (MSE) ➘ Human Visual System (HVS) based metrics

✔Blind Measurement

➘ Original reference images are not available

SLIDE 4

Proposed Measurement System

Differen

tiation

| . | Power Spectrum Estimation Bicoherence Estimation Vertical Blocking Measure Differen

tiation

| . | Power Spectrum Estimation Bicoherence Estimation test image rows columns Weighting & Summing weights weights blocking measure FFT FFT Horizontal Blocking Measure

SLIDE 5

Power Spectrum

✔ Comparison of power spectra of the blocky and the

riginal images

✔ The blockiness is characterized by the peaks at several

feature frequencies

0.1 0.2 0.3 0.4 0.5 0.5 1 1.5 0.1 0.2 0.3 0.4 0.5 0.5 1 1.5

SLIDE 6

Power Spectrum (continued)

✔ In some cases, the image signal itself has a special

frequency distribution that may disturb the characteristic frequency components.

✔ Difficult to get a robust measure on power spectrum

0.1 0.2 0.3 0.4 0.5 0.5 1 1.5 0.1 0.2 0.3 0.4 0.5 0.5 1 1.5

SLIDE 7

Ideal 1-D ‘blocky’ signal

✔ 1-D blocky signal

5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0 0 . 2 0 . 4 0 . 6 0 . 8 1

✔ Being differentiated and applied absolute operator

5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0 0 . 0 1 0 . 0 2 0 . 0 3 0 . 0 4 0 . 0 5 0 . 0 6

SLIDE 8

Idea 1-D ‘blocky’ signal (cont.)

✔ Phase of FFT result

5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0

1 2 3 4

✔ Magnitude of FFT result

5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0 0 . 5 1 1 . 5 2 2 . 5 3 3 . 5 4

SLIDE 9

Bispectrum - Blocky Image

SLIDE 10

Bispectrum - Original Image

SLIDE 11

Blocking Metric

✔Vertical Blocking Metric

]) 8 / 3 [ ] 4 / [ ] 8 / [ ( ] 4 / , 8 / [

N P N P N P N N C M

XXX Bv

+ + ⋅ ⋅ = γ

✔Overall Blocking Metric

Bh Bv B

M M M 5 . 5 . + =

SLIDE 12

Modified Measurement System

Differen

tiation

| . | Power Spectrum Estimation Bicoherence Estimation Vertical Blocking Measure Differen

tiation

| . | Power Spectrum Estimation Bicoherence Estimation test image rows columns Weighting & Summing weights weights blocking measure FFT FFT Horizontal Blocking Measure Masking Masking Masker Evaluation

SLIDE 13

Masker Evaluation

✔ Luminance masking

➘ more sensitive to mid-

level errors

✔ Local activity masking

➘ more sensitive to errors in

smooth areas

Brighter - stronger masker Darker - weaker masker

SLIDE 14

Measurement Results - ‘Lena’

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 5 10 15 20 Proposed blocking measure Combining masking effects

SLIDE 15

Measurement Results - ‘Barbara’