Digital Image Tamper Detection Based on Multimodal Fusion of Residue Features - Advanced Concepts for Intelligent Vision Systems - page 83

Table 1. Evaluation of noise and quantization residue features for emulated copy-move tamper

~

f f

attack (% Accuracy); ( noise residue features); (quantization residue

Intra − Inter Intra − Inter

features)

Internet movie data subset % Accuracy

Different Residue features and their fusion

CCA ICA FLD ARC[10]

83.2 83.4 83.6 80.2

f

(Intra-frame noise residue features)

Intra

83.8 83.1 83.4 83.1

f

(Inter-frame noise residue features)

Inter

~

77.28 80.26 76.23 74.33

f

(Intra-frame quant. residue features)

Intra

~

72.65 78.27 71.44 69.45

f

(Inter-frame quant. residue features)

Inter

86.6 86.1 85.27 83.78

f

(feature fusion- noise residue)

Intra Inter

~

80.55 82.34 79.66 77.22

f

feature fusion- quant residue)

Intra Inter

~

89.56 88.85 86.22 84.33

f f

+ (hybrid fusion)

Intra Inter Intra Inter

Table 2. (% Accuracy) Performance for noise and quantization residue features and their fusion
for GMM vs. NN classifier

% Accuracy GMM

Classifier

SVM
Classifier

NN
Classifier [10[

Different Residue features and their fusion CCA features CCA features CCA features

83.2 83.4 81.4

f

(Intra-frame noise residue features)

Intra

83.8 83.5 80.6

f

(Inter-frame noise residue features)

Inter

~

77.28 78.18 75.77

f (Intra-frame quant. residue features)

Intra

~

72.65 74.43 70.53

f

(Inter-frame quant. residue features)

Inter

86.6 84.96 83.22

f (feature fusion- noise residue)

Intra Inter

~

80.55 82.43 77.23

f

feature fusion- quant residue)

Intra Inter

~

89.56 90.56 83.45

f f

+ (hybrid fusion)

Intra Inter Intra Inter