Skip to main content

Table 5 Overall accuracy in noisy environment

From: Multilayer graph spectral analysis for hyperspectral images

(Uniform/Gaussian)

Indian Pines

Salinas

TS/C

Noise level

MPCA

MLG-MRC

MSPCA

MLG-MRC

Setup 1

     

5

5%

0.7140/0.7012

0.7377/0.7493

0.8592/0.8670

0.9231/0.9159

10%

0.7110/0.6702

0.7135/0.7276

0.8341/0.8421

0.8956/0.9023

15%

0.6868/0.6507

0.7133/0.7215

0.8107/0.8215

0.8815/0.8975

5 10

5%

0.7613/0.7753

0.8033/0.8056

0.9509/0.9647

0.9632/0.9684

10%

0.7522/0.7498

0.7788/0.7984

0.9408/0.9431

0.9445/0.9491

15%

0.7317/0.7382

0.7544/0.7845

0.9381/0.9368

0.9421/0.9317

Setup 2

     

5 5

5%

0.6839/0.6740

0.7023/0.6898

0.8591/0.8679

0.9240/0.9050

10%

0.6640/0.6501

0.6764/0.6647

0.8471/0.8512

0.9045/0.8878

15%

0.6413/0.6247

0.6687/0.6427

0.8317/0.8421

0.8915/0.8775

5 10

5%

0.7579/0.7408

0.8022/0.7810

0.9682/0.9621

0.9735/0.9633

10%

0.7409/0.7333

0.7800/0.7762

0.9521/0.9450

0.9547/0.9532

15%

0.7208/0.7235

0.7695/0.7584

0.9437/0.9416

0.9305/0.9241

  1. The best performances are marked in bold