Image EnhancementDenoising & Contrast Enhancement Mingzhu Long - - PowerPoint PPT Presentation

▶

Feb 05, 2024 446 likes •686 views

Image EnhancementDenoising & Contrast Enhancement Mingzhu Long 2016.02.05 MENU n Image Denoising Algorithm Classification Examples n Image Contrast Enhancement Algorithm Classification Examples n Image Quality Assessment MENU

SLIDE 1

Image Enhancement—Denoising & Contrast Enhancement

Mingzhu Long 2016.02.05

SLIDE 2

Image Denoising

ØSingle Image Denoising

ØSpatial Domain

ØLocal ØNonlocal

ØTransform Domain

ØData adaptive transform ØNon-data adaptive transform

from spatial domain to transform domain, from local to nonlocal methods, from pointwise to multipoint methods, from single image to multi-image.

ØMulti-image Denoising

Ømotion compensation ØNon-motion compensation

SLIDE 5

Examples--NLM

A nonlocal algorithm for image denoising. CVPR 2005.

The NL-means not only compares the grey level in a single point but also the geometrical configuration in a whole Neighborhood. This fact allows a more robust comparison than neighborhood filters.

SLIDE 6

Examples--NLM

A nonlocal algorithm for image denoising. CVPR 2005.

Noisy image(standard deviation 20)，Gauss filtering, anisotropic filtering, Total variation, Neighborhood filtering and NLmeans algorithm

SLIDE 7

Examples--BM3D

Image Denoising by Sparse 3-D Transform-Domain Collaborative Filtering. TIP 2007.

Step 1)Basic estimate. a) Block-wise estimates. Grouping. Collaborative hard-thresholding. b) Aggregation.

时，效果明显下降。计算量，存储。

Step 2)Final estimate: a) Block-wise estimates. Grouping. Collaborative Wiener filtering. b) Aggregation.

时，效果明显下降。计算量，存储。

SLIDE 8

Examples--BM3D

Image Denoising by Sparse 3-D Transform-Domain Collaborative Filtering. TIP 2007.

SLIDE 9

Examples--Multi-image

Multiple view image denoising. CVPR 2009.

How to deciding whether a patch A1 is similar to a patch B1 in the reference image I1? We find their corresponding patches A2 and B2 ,respectively, in the second image I2. If A1 is similar to B1, A2should also be similar to B2.

SLIDE 10

Examples--Multi-image

Multiple view image denoising. CVPR 2009.

Noisy patch BM3D Our method Ground truth

SLIDE 11

Image Contrast Enhancement

Intensity Transformation

Logarithmic transformation Gamma transformation Contrast stretching

Histogram Equalization

Global histogram equalization Local histogram equalization

‹ 8/16

SLIDE 13

Results-Log

9/16

SLIDE 14

Results-Gamma

Gamma correction depends on a suitable selection of the index. It seems that one index can not take into account both the details in the dark area and lighter area. 10/16

SLIDE 15

Results-HE

11/16

SLIDE 16

Results-HELoc

12/16

SLIDE 17

Results

It seems that CE&NLM is clearer than NLM&CE The histogram makes some areas too bright enhancement 6/16

SLIDE 18

Solutions

nCombination of different methods[1] or index[2]!

ØBrightness estimation ØThreshold estimation

nTransformation domain contrast enhancement？

[1] Gabriel Zahi. Automatic Detection of Low Light Imagesina Video Sequence Shot under Different Light

Conditions. 2015.

[2] Jinfang Shi. A Novel Image Enhancement Method Using Local Gamma Correction with Three-level

Thresholding. 2011

13/16

SLIDE 19

Image Quality Assessment

nSubjective Evaluation nObjective Evaluation

ØFull-reference Evaluation Method

ØPSNR, MSE, SSIM, Entropy，etc.

ØSemi-reference Evaluation Method ØNon-reference Evaluation Method

ØEvaluation methods for some certain distortions ØGeneral evaluation method

SLIDE 21

图像质量评价

nMean Square Error nNormalized Mean Square Error nMean Absolute Error

SLIDE 22

图像质量评价

nNormalized Mean Absolute Error nPeak Signal-to-Noise Ratio nStructural Similarity

SLIDE 23

THANKS！ Q&A

Mingzhu Long 2016.02.05