Corn (Zea Mays) is one of the most sensitive crops to plant arrangement and plant density. The most commonly used method to scout plant density is by visual inspection on the ground. This field activity becomes time consuming, observation biased, and may lead to less-profitable decisions by farmers. The objective of this study is to develop a method for plant count estimation based on high resolution imagery taken from UAS at low altitude with application to monitor early season crop performance at field conditions. A Sony ILCE α5100L RGB camera with 24 Mpixels with a 16-50 mm focal length lens was flying aboard an octocopter. A ground sampling distance of 2.4 mm is targeted to extract information at a plant level basis. First, an ExG transformation is used to separate green pixels from the background, rows and interrow contours are then identified and extracted. A scalable training procedure is implemented using geometrical descriptors as inputs of the classifier. Second, a decision tree is implemented and tested using two training modes and degraded resolution simulated data in each site to expose the workflow to different environmental and management conditions at the time of the aerial data acquisition. Differences in performance were due to training modes and spatial resolutions in the two sites. An accuracy of 0.96 and 0.93 were found for local and joined training modes, respectively. Plants should have between two to three leaves when images are taken. The best performance of the workflow is reached at 2.4 mm resolution corresponding to 10 meters altitude. It is coincident with the larger number of green objected detected in the images and the effectiveness of geometry as descriptor for corn plant detection.