In the process of economic development of a country, agriculture plays a strategic role. For precision agriculture science, data stability and accuracy of the rice seedlings are important issues in estimating rice productions for paddies. With the rapid development of wireless communication systems and aerial photography techniques, drones can provide essential full-color images for the applications of rice paddy fields. A proper cultivation density of rice seedlings is important for agricultural data. Thus, this article presents the approaches for the detection and interpretation of transplanted positions in terms of drone’s eye-view images for rice paddies. In the light of deep learning algorithms, the proposed approaches built a model for estimating the transplanted positions of paddies from the type of training set. The validation was employed to obtain an 80/20 dividing ratio of training/test data, and then the set of well-known coordinates of the seedlings of rice paddies is used to validate the efficiency. The results demonstrate that the accuracy of the detection and interpretation of transplanted positions for rice seedlings is higher than 94.88% in terms of the F 1 -measure value.