Several earth observation satellites bear optical imaging sensors whose outputs are essential in many environmental aspects. This paper focuses on two parameters of the quality of the imaging system: the Modulation Transfer Function (MTF) and Signal to Noise Ratio (SNR). These two parameters evolve in time and should be periodically monitored in-flight to control the quality of delivered images and possibly mitigate defaults. Only a very limited number of past and current sensors have an on-board calibration device fully appropriate to the assessment of the noise and none of them has capabilities for MTF assessment. Most often, vicarious techniques should be employed which are based on the Earth-viewing approach: an image, or a combination of images, is selected because the landscape offers certain properties, e.g., well-marked contrast or on the contrary, spatial homogeneity, whose knowledge or modeling permit the assessment of these parameters. Several methods have been proposed to perform in-flight assessments. This paper proposes a review of the principles and techniques employed in this domain.
