In order to retrieve the defocus in the optical system of airborne camera caused by the changes of temperature, pressure and other environment parameters in high altitude, focus measure algorithm based on power spectra and auto-focus search strategy are represented. Firstly, it is introduced that the power spectra of images taken from different region has the characteristic of invariance. Two betterments over the power spectra sum (PSS) focus measure function are proposed based on the main attenuation of power spectra in the high frequency. Then, full search (FS) strategy using changed steps is proposed in order to reduce the focus error caused by noise from environment. The results show that the FS strategy can locate the optimal focus position accurately when using search step length of 1 mm. The focus error is 0.3 mm which fulfills the requirement of system. The PSS proposed algorithms can indicate the defocus state of images taken by airborne camera, and it can also improve the properfies of saturation region and sensitiveness of the focus measure curve respectively. Search steps shorter than focus-depth can guarantee the validity of the result.