An Image Fusion Algorithm Based on Discrete Wavelet Transform and Canny Operator - Advanced Research on Computer Education, Simulation and Modeling - page 36

(a) Original image 1 (b) Original image 2

Fig. 2. Original images

(a) Method of average

(b) Method of this article 1

(c) Method of this article 2

(d) Method of this article 3

Fig. 3. Comparison of four fused images

Qualitative analysis displays that the visual effects of the latter two fused images are
better than that of the first two images. Then make a quantitative analysis from three
respects: independent factors (IF), united factors (UF) as well as comprehensive
evaluation (CE)[9] . Mean value (MV) reflects average brightness as to human eyes.
Mean gradient (MV) indicates the clarity, detail contrast and texture transformation
features. Information entropy (IE) can show the quantity of image information. De-
viation index (DI) reflects the deviation between fused images and reference image and
the smaller its value is, the closer two images' similarity is[10].