A new descriptor for UAV images mapping by applying discrete local radon

It has been proven that feature based image descriptors are successful in recent years. However, a light and powerful feature descriptor is still needed to match images in real time. In this paper, we presented a new discrete descriptor for areal images that are obtained from UAVs-mapping to increase the speed and accuracy of output ortho-map. This binary descriptor takes the advantage of the local Radon descriptors to create the binary codes. Our preliminary investigations have shown the local Radon transformation can be an effective method to apply in key-point descriptors, especially in aerial image mapping.