| Summary: | The shape and area of the vortex are important indicators to judge the effect of chemical liquid spraying and crop pollination by agricultural plant protection UAV. However, the traditional way of extracting the vortex related data through manual methods without an adaptive mechanism is complicated and inefficient. In order to solve this problem, the vortex image is preprocessed by adaptive iterative binarization and mean filtering. Then, the Convex Hull algorithm combined with Edge function is applied to the pre-processed image to extract the specific shape of vortex image and further calculate its area. Experimental results show that this method can be used to extract the vortex shapes and calculate the vortex area under different experimental conditions without manual intervention. Compared with the results of vortex area obtained via the manual extraction method, this method reported the average absolute error rate of 2.84%, the RMSE of 0.2903, and the correlation coefficient of 0.9965, indicating that the proposed method is more applicable to reveal the actual state of the vortex. A standardized extraction method was established to provide a standard for future vortex image research. The method in this paper lays a scientific foundation for the real-time tracking research of agricultural plant protection UAV and vortex.
|
|---|
