UAV-Based Identification of Achnatherum Splendens Community Combining K-Means and Artificial Fish Swarm Algorithm

The Achnatherum splendens community is one of the important types of steppe vegetation, which plays a key important role in environmental protection. A study on the unmanned aerial vehicle (UAV) based image segmentation was carried out to identify the A. splendens community in the Hulunbeier Pasture of Inner Mongolia, China in May 2016. Three typical types of A. splendens community were categorized including aggregated type, dispersed type and mixed type. K-means clustering and artificial fish swarm algorithm were jointly used to improve the accuracy. The segmentation effect was estimated by the measured values and the calculated values derived from the Can-Eye software. The results showed that the difference between the measured values and KM-AFSA was smaller than that between the measured values and Can-Eye software. In addition, the coefficient of determination (R²) could reach 0.98 between the measured values and KM-AFSA. The study could provide a reference for the identification of other grassland vegetation using UAV-based photographs.