An application of the modern methods for satellite image processing for solution of problems of environmental monitoring

• Дата: 2024
• Автор: Alpert, Sofiia
• Формат: Стаття
• Мова: Ukrainian
• Опубліковано: Scientific Centre for Aerospace Research of the Earth Institute of Geological Science National Academy of Sciences of Ukraine, Kyiv, Ukraine 2024
• Теми: image classification; Normalized Difference Vegetation Index; evidence theory; Yager’s combination rule; ecological monitoring
• Онлайн доступ: https://ujrs.org.ua/ujrs/article/view/260
• Назва журналу: Ukrainian Journal of Remote Sensing of the Earth

Modern Remote Sensing methods and approaches gives a new opportunities for conducting scientific research in a much more detail way. Nowadays many methods for satellite image processing are applied in remote sensing. Band selection and classification procedure are one of most serious and complex procedures of satellite image processing. Band selection method based on correlation analysis, Yager’s combination rule and Normalized Difference Vegetation Index (NDVI) for satellite data processing were proposed in this paper. Application of NDVI is the first step of classification. The NDVI can be applied to evaluate the density of green vegetation. Different values of Vegetation Index correspond to different classes of objects, such as: sand, soil, reservoirs, green vegetation, roads and petroleum pollutions. Using NDVI, we can select special classes, that we need. Control classification is the second step of satellite image processing. But a lot of classification methods can not deal with conflicting data and can provide illogical and wrong results of classification. That’s why we should use the Dempster-Shafer evidence theory and Yager’s combination rule. Yager’s rule can process ambiguous and incomplete data from different spectral bands. Main advantages of the Dempster-Shafer evidence theory and Yager’s combination rule were described and analyzed in this work. The development of Dempster-Shafer evidence theory arises from the necessity to overcome the limitations of Probability Theory. It was noted, that Yager’s combination rule can quickly and easy process information. It was given a formula of Yager’s combination rule in this paper. Yager’s rule can combine imprecise and high conflicting information. It was considered an numerical example, where NDVI and Yager’s combination rule were used for detection and mapping of petroleum pollution. Described methods for satellite image processing can be applied in different agriculture and practical tasks, searching for minerals, ecological monitoring, mapping of petroleum pollutions.