Table 3 Comparison of results of various infrared image enhancement methods.

From: Dual decoding generative adversarial networks for infrared image enhancement

 

Method

OG

HE

CLAHE

GF

SSR

MSR

TIECNN

FUnIE

Ours

Img1

AG

6.0248

21.5455

20.3509

10.9852

7.8593

7.9757

19.6100

21.7028

21.3267

IE

6.2390

7.9608

7.7017

6.6686

6.6707

7.7514

7.8571

7.9113

7.9210

PSNR

7.8692

23.3966

14.2653

8.2680

8.8879

9.1173

21.3133

25.0579

25.5274

SSIM

0.3474

0.8459

0.7579

0.3994

0.4578

0.4520

0.8584

0.9166

0.9258

Img2

AG

5.1035

21.4965

18.1191

7.0424

12.0773

10.4535

14.2562

16.0262

17.1275

IE

5.8133

7.8804

7.1668

5.9153

7.7584

7.7491

7.4697

7.5638

7.7395

PSNR

10.6677

14.9974

16.1804

10.4561

15.0742

15.1718

21.5889

23.6008

26.9968

SSIM

0.4025

0.6304

0.7429

0.3764

0.6604

0.6363

0.7852

0.8452

0.8953

Img3

AG

4.0684

27.9084

13.6946

6.4984

10.3271

9.9790

12.4966

13.4100

15.8442

IE

5.0308

7.1811

6.8459

5.2812

6.9053

6.9398

6.6147

6.7015

7.0372

PSNR

15.3018

9.1228

23.4671

15.9015

8.8818

9.1681

23.3151

24.0755

27.2207

SSIM

0.4728

0.3746

0.7807

0.4856

0.4685

0.4754

0.8177

0.8379

0.8839

  1. Best values for each metric are in bold.