Mean Shift Clustering with Bandwidth Estimation and Color Extraction Module Used in Forest Segmentation

This study aims to implement a useful mean shift clustering method to identify forest areas on aerial images due to the need to combat and monitor deforestation worldwide. Two challenges have been addressed in the process: estimating the best algorithm bandwidth and friendly representation of the results. Also, a comparison has been done with previously implemented algorithms using the final computed accuracies.