Skip to main content

Table 6 Performance comparison without rotation for each test carried out to assess the robustness

From: A hybrid method for accurate iris segmentation on at-a-distance visible-wavelength images

Blur

Noise

Dice

Precision

Recall

Error

Exec. time

0

-

0.9121

0.8972

0.9322

0.0102

0.9187

0

Gaussian

0.8904

0.8552

0.9336

0.0153

0.7488

0

Poisson

0.9095

0.8904

0.9342

0.0121

0.7827

0

Salt & pepper

0.8676

0.8405

0.9157

0.0166

0.7805

3

-

0.8934

0.9061

0.8886

0.0141

0.9614

3

Gaussian

0.8782

0.8551

0.9086

0.0169

0.7489

3

Poisson

0.8920

0.8984

0.8924

0.0143

0.7676

3

Salt & pepper

0.8414

0.8341

0.8729

0.0192

0.8245

6

-

0.7805

0.8374

0.7462

0.0258

0.8850

6

Gaussian

0.8093

0.8064

0.8260

0.0259

0.7472

6

Poisson

0.8127

0.8512

0.7911

0.0227

0.7358

6

Salt & pepper

0.7494

0.7687

0.7716

0.0275

0.7896

  1. Segmentation evaluation and execution time (s) per image