This paper presents a correction technique for the geometric distortion that affects images from cameras aboard UAV when in off-nadir conditions. This kind of lightweight platform is extremely susceptible to turbulences, and often does not have a gyro stabilized gimbal to keep camera pointed to nadir, so it is common to occur off-nadir extrapolations with significant effects on the image geometry. So, when you want to extract reliable metric information from the image or to enable accurate matching in a Geographic Information System – GIS in order to perform fusion or geoprocessing, this distortion must be known and corrected. Among the various existing correction methods, we present here the parametric approach that uses the collinearity equations, that requires camera external orientation parameters. We present all mathematical formulations for the correction, definitions of camera geometric parameters and how to calculate them. Another important contribution of this paper is the demonstration of difference between the roll-pitch-yaw angles commonly used in aeronautical systems, and the omega-phi-kappa angles commonly used in photogrammetric systems. People may make some confusion about that when extracting UAV data to be used in photogrammetry software. We present clear definition of reference axes and rotation angles for each system and the mathematical development to perform conversion. For practical demonstration, we applied these techniques to some off-nadir images from a small-format camera on a fixed-wing UAV with known position and attitude angles and. Thus, we were able to perform image resampling to correct the projective distortion.