Hyperspectral sensors provide 3-D images with high spatial and spectral resolution. Acquired data can be utilized in diverse applications such as detection and control of hazardous agents in atmosphere and water, military targets, and so on. Over the last decade, hyperspectral remote sensing algorithms for target detection have evolved from the spectral-based methods, which only use spectral information, to more recent methods based on spatial-spectral information. Spatial information plays a crucial role to improve the efficiency of the algorithms. Furthermore, the parallelization of the algorithms reduces the computation time. Developments in the area of commodity computing provide affordable approach for target detection applications with real-time constraint. We will give a scientific overview of recent target detection algorithms which try to overcome existing limitations (e.g. spectral variability or background interference) in hyperspectral remote sensing. Unlike current target detection methods in literature, this study explains and assesses different aspects of developments in target detection algorithms comprehensively. In particular, this study focuses on development in atmospheric correction methods which especially deal with background interference, development in methods based on spectral information and spectral-spatial information (both methods especially deal with spectral variability), and parallelization of the algorithms. With consideration of hyperspectral data challenges in real-world, an optimum approach is the adaptive algorithm based on spatial-spectral information in which their computation is performed in parallel