Aiming at the problem that traditional infrared DOAS （Differential Optical Absorption Spectroscopy） cannot retrieve low-resolution spectrum and the high cost of the high-resolution spectrometer， this study investigates and develops an envelope retrieval algorithm applicable to low-resolution spectral retrieval. By examining the shape of absorption cross-sections and spectral broadening， considering the retrieval of low-resolution spectra， and integrating with the infrared DOAS algorithm， an envelope retrieval algorithm capable of retrieving atmospheric CO? concentrations using low-resolution spectrometers is developed. Integrated with the near-infrared hyperspectral remote sensing instrument， relevant detection experiments were carried out on Science Island in Hefei. Using the LBLRTM radiative transfer model and the HITRAN database， a series of retrieval concentration data were obtained and analyzed. The results indicate that the traditional infrared DOAS algorithm failed to perform normal retrieval， whereas the envelope retrieval algorithm can be applied to low-resolution spectrometers， with the retrieval residual below 1.05% and a correlation coefficient of 0.76 with the TCCON （Total Carbon Column Observing Network） site data. This study proposes an algorithm applicable to low-resolution spectral retrieval， breaks through the limitations of the traditional infrared DOAS algorithm， and provides a new method for achieving low-cost and accurate CO? detection.