Aiming at the problem of anti bait interference in infrared precise guidance, the simulation and experimental research of target spectrum recognition technology based on spectral angle and spectral distance were carried out. On the premise that target and bait were both gray body radiation, the infrared radiation distribution of target and bait under different radiation temperature and different spectral emissivity models was modeled. The relationship between temperature, spectral emissivity and spectral angle, spectral distance of target and bait was quantitatively analyzed, and the spectral difference between target and bait was numerically characterized. In order to verify the effectiveness of the spectral discrimination method, the target and bait spectral data were collected and analyzed. The results show that the spectral angle and spectral distance can be used to distinguish target from bait, which can provides theoretical and experimental basis for target classification in complex background in the future.