In recent years, the emergence of Unmanned Aerial Vehicles (UAV), also known as drones, with high spatial resolution, has broadened the application of remote sensing in agriculture. However, UAV images commonly have specific problems with missing areas due to drone flight restrictions. Data mining techniques for imputing missing data is an activity often demanded in several fields of science. In this context, this research used the same approach to predict missing parts on orthomosaics obtained by UAV using a PlanetScope and Sentinel-2 images as auxiliary data. The spectral bands (blue, green, red, red-edge and infrared) and the NDVI (Normalized Difference Vegetation Index) derived from the satellites images were used as predictor variables. The study area is located in Caiuá municipality, western of São Paulo State, Brazil. The imaged area covered 200 ha of pasture cultivated under an integrated crop-livestock system. The sensor boarded in the UAV was a multspectral camera Micasense RedEdge (RedEdge TM, Micasense, Seattle, Washington, USA). UAV images were acquired using a quadrotor drone, which performed the flight plan automatically. The established flight height was 115 m with 0.08 m of GSD (Ground Sample Distance) and overlap equal to 75 %, both sides. The flight window was between 9.30 a.m. and 1.30 p.m. The Sentinel’s image acquired on August/11^th and PlanetScope on August/10^th. Likewise, UAV data collection mismatched a maximum of 1 day to satellites. Because platforms have different spatial resolutions, we first resampled the UAV (0.08 m), PlanetScope (3 m) and Sentinel-2 to 1 m, so that all of the images have the same spatial resolution. We tested different proportions of overlapping parts in the UAV and satellites images to train the machine learning algorithm Random Forest. The absolute and relative Root Mean Square Errors (RMSE) and coefficient of determination (R²) were calculated to assess the accuracy of the machine learning models. The prediction accuracies of the machine learning models were little influenced by the missing parts. The association of the high spatial resolution of PS images and the high spectral resolution of S2 images results in improved performance of the RF-based imputation models. This framework using Random Forest and the PlanetScope and Sentinel images may be applied to fill the gaps of UAV orthomosaics. Data mining techniques are feasible for the proposed objective, being necessary the adequate parameters definition that allow reaching the best prediction quality of the models to fill in the missing areas in UAV images.