27.12.2023Mapping of Vegetations through Machine Learning Algorithms

My internship at the Philippine Space Agency has broadened my knowledge in remote sensing and vegetation mapping, offering a combination of learning and professional growth. I was immersed in the practical applications of optical images and learned the complexities of using machine learning algorithms such as Random Forest and Support Vector Machines which are frequently used for classifying and mapping vegetation classes with high accuracy.

Working in an environment such as in a Space Agency is a core for innovation and collaboration. My superiors and colleagues were supportive, sharing valuable insights and suggestions like using NDVI for image masking and the use of principal component analysis for enhancing machine learning classifications. They also guided me in the nuances of classifying thin canopy and grass, which is a common challenge in using these images.

One of the highlights of my experience was the hands-on engagement with various machine learning algorithms using different GIS software like SNAP, QGIS, ENVI, and ArcGIS. The study that was conducted was challenging but rewarding. While working with the different classification techniques, I also faced multiple hurdles like software glitches and unfavourable results which taught me the importance of flexibility and adaptability.

Organisation of the internship

For those who are considering this internship, a background in remote sensing techniques and basic programming can be helpful. Prospective interns should be aware that the living and housing expenses are not covered by the program. If this opportunity aligns with your academic goals, I encourage you to reach out through their website. This internship promises you a rich learning experience in space technology. 

Lou Lerren Chan Curacha