PROJECT TITLE :

DenseFuse A Fusion Approach to Infrared and Visible Images

ABSTRACT:

For infrared and visible image fusion challenges, we have developed a revolutionary Deep Learning architecture. With our encoder, the outputs of each layer are connected to the outputs of all the other layers, unlike standard convolutional networks that use only convolutional layers and a fusion layer. Two fusion layers (fusion methods) have been devised as part of this architecture in order to combine the beneficial information extracted from source images during encoding. Finally, a decoder reconstructs the merged image. The suggested fusion method outperforms existing fusion methods in both objective and subjective assessment.