Fusion of Night Vision and Thermal Images

摘要

Night vision and thermal images are extensively used in military operations, as they help in mission planning and executions tasks. Image fusion effectively combines information present in each type of image. This research explored two wavelet-based image fusion approaches for night vision and thermal images; namely wavelet transform fusion and region-based fusion. Morphological methods designed to improve the image segmentation step were considered to improve image contrast and a global image quality index was applied to investigate the information content improvement resulting form the fusion process. Finally, a MATLAB-based graphical user interface was designed to assist the user in evaluating the benefits of the fusion process. Results showed the selection process is able to narrow to the best fused image with a satisfactory accuracy.