Artificial intelligence researchers are very interested in computer vision in drones. Many real-time problems will be solved by providing intelligence to drones. Object identification, tracking, and counting are all important computer vision tasks for monitoring certain situations. However, altitude, camera angle, occlusion, and motion blur make it a more difficult endeavour. Methodology: A thorough literature evaluation concentrating on object recognition and tracking utilising UAVs in various applications has been undertaken in this work. This study reviews previous research publications conclusions and indicates research needs. The object detection algorithms used in UAV photos are categorised and further upon. There is a collection of UAV datasets for object detection tasks. Existing research studies in various applications have been summarised. Finally, in precision agriculture, a secure on board processing system based on a strong object detection framework is presented to address identified research needs.