국토지리학회지 ISSN:1225-3766(Print) THE GEOGRAPHICAL JOURNAL OF KOREA

In this study, we set up survey points at elevations of 1,500 m, 1,600 m, and 1,700 m on Mt. Halla to investigate the seasonal biological characteristics of indigenous plants in response to climate change using high-resolution drone imagery. We conducted drone surveys a total of seven times over two years: three times in May, August, and October in 2021, and four times in March, May, August, and October in 2022. Based on the survey results from the first year (2021), we used both pixel-based and object-based classification methods to classify the indigenous plant species and communities in the target area. We verified that the object-based classification method was valid for high-resolution drone images. Through the results of the second year (2022) surveys, we were able to perform high-resolution mapping and analysis of vegetation indices at the survey points. Moreover, we developed and applied several workflows to improve the object-based classification method for classifying indigenous plant species and communities, and we were able to present the optimal classification model. By employing this newly developed object-based classification method, we achieved classification accuracies of 90.67% at Point 1, 90.00% at Point 2, and 84.67% at Point 3, respectively.