Atmosphere is an important factor that affects the quantitative analysis and application of remote sensing technology. In the support of IDL platform, the study takes advantages of dark pixel atmospheric correction algorithm (DPACA) to extract the optical depth of atmosphere, and then remove the atmospheric influences from each channel of the Hyperion sensor by that atmospheric parameter. The study results show that the optical depth decreases with the increase in central wavelength of the Hyperion sensor, namely the optical depth is negative versus the central wavelength of sensor. The linear model is the optimal experimental model that is used to describe that relationship, and its correlative coefficient is 0.912 3. It was found that the signals recorded by the remote sensing sensor can’t express the inherent optical properties and apparent optical properties in a proper manner. Additionally, the remote sensing signals are insensitive to the variations of waters qualities’ samples. At the blue and green bands, the effects of atmosphere are the most serious. The spectra are completely different from the optical properties of natural waters at those bands. Compared with the theoretical spectral features of waters’ optical properties, the image quality of Hyperion sensor has been perfectly improved by the DPACA. Under the condition of lacking the vertical profile data of atmosphere, the DPACA is an available approach to removing the atmospheric influence on the Hyperion imageries.