In last decades the hyperspectral (HS) imaging have been used vastly, as it provides very good temporal, spatial, and spectral resolutions of larger areas. The hyperspectral imaging provides accurate estimation similar objects with tolerable spectral signature and accurate estimation of complex surface physical properties; thus, HS imaging have been used in variety of remote sensing application such as military, crop monitoring, environmental study, and precision agriculture etc. The fusion of two images i.e., one with higher spectral information and other with higher spatial information provides huge scope in aforementioned application. This paper provides an extensive survey of various image fusion methodologies using statistical, machine learning (ML), and Deep learning methodologies for hyperspectral images. The survey shows the traditional panchromatic and multispectral (MS) image fusion technique requires high precision registration, which impose limitations in achieving high-quality fusion outcomes. Thus, fusion of MS and HS have been used in recent time for provisioning different remote-sensing application; thus, it prerequisite to minimize uncertainties for achieving better fusion enhancement. This paper provides present methodologies, problems, and possible solution of hyperspectral image fusion for remote sensing application.