Research Article

Adaptive Compressive Sensing of Images Using Spatial Entropy

Table 1

PSNR (dB) comparison of various CS-based codec for test images at different measurement rates.

Test imageBCSV-ABCS [6]S-ABCS [7]Proposed
PSNRΔPSNRPSNRΔPSNRPSNRΔPSNRPSNR

= 0.1
Lenna18.89−8.1918.58−8.5019.72−7.3627.08
Barbara16.64−5.2917.02−4.9117.05−4.8821.93
Peppers17.28−9.1017.46−8.9218.31−8.0726.38
Goldhill20.49−5.5920.47−5.6121.24−4.8426.08
Mandrill15.71−3.9318.14−1.5018.89−0.7519.64
Avg.17.80−6.4218.33−5.8919.04−5.1824.22

= 0.3
Lenna27.35−5.3429.05−3.6430.54−2.1532.69
Barbara24.02−0.8125.600.7725.941.1124.83
Peppers24.35−6.9628.59−2.7229.34−1.9731.31
Goldhill23.86−6.4926.70−3.6527.08−3.2730.35
Mandrill17.64−5.1319.74−3.0319.39−3.3822.77
Avg.23.44−4.9525.94−2.4526.46−1.9328.39

= 0.5
Lenna31.64−4.632.10−4.1434.41−1.8336.24
Barbara28.350.7029.271.6230.693.0427.65
Peppers31.11−3.0131.18−2.9432.70−1.4234.12
Goldhill29.19−4.129.19−4.130.61−2.6833.29
Mandrill21.04−4.3122.76−2.5922.82−2.5325.35
Avg.28.27−3.0628.90−2.4330.25−1.0831.33