Satellite remote sensing, which has the ability of searching large areas, plays an important role in the emergency rescue of natural disasters. In this paper, based on the spectral characteristics of different substances, a multi-color and multi feature target identification measurement for target detection is proposed. Firstly, the optimized selection is carried out the spectral section selection, which reduces the data volume of the system and improves the signal-to-noise ratio of the system. Secondly, in terms of information acquisition, multi-dimensional information (space, spectrum, motion, radiation and background) of spatial moving targets can be detected comprehensively through satellite motion, and the recognition probability of detection can be improved through multi-color and multi-feature comparison. At the same time, using the bifocal length imaging system, using imaging system, using the long focal length and short focal length two optical channel common aperture technology, through the reasonable choice of system structure and optimization design method, reduce the system volume, reduce the weight of the system, so as to achieve based on spectrum- space multiplexing bifocal length imaging system miniaturization design. Finally, the new combined detection mode is combined with parallel processing algorithm to realize fast search and recognition.