Image enhancement based on quality pre-assessment and weighted wavelet fusion for WSN nodes

According to the characteristics of wireless sensor networks, image enhancement for WSN nodes is presented in this paper. In particular, the source images are pre-evaluated and unqualified images are rejected; then, qualified images are decomposed into several coefficients by using wavelet transform; finally, the fusion image is obtained by weighted wavelet fusion. Experimental results, using four sample images, show that the decision made by majority voting correlate highly with subjective judgment, and data volume has reduced up to 75%. These results compare very favorably with the results obtained with other methods.