An Underwater Image Mosaicing and Fusion Approach based on Weighted Aggregation Energy Threshold using Multi-wavelet Transform

As ideal tools, autonomous underwater vehicles (AUVs) are used to implement underwater monitoring missions instead of human. Because the underwater images captured are many close-range images, the mosaicing and fusion method is applicable to create large visual representations of the sea floor. A novel mosaicing and fusion approach based on weighted aggregation energy threshold using Biorthogonal wavelet transform was proposed in order to improve the quality and contrast of the underwater images. Firstly, adopting the phase correlation method, the overlapped areas of the source images were determined, and then the overlapped areas were decomposed using Biorthogonal wavelet transform. Secondly, the overlapped area were fused and constructed adopting the low-frequency and high-frequency images fusion algorithms, and the mosaicing image with clarity contrast was obtained. Finally, fusion performance evaluations were carried out to evaluate the qualitative and quantitative of the image.