Bøger af Prasadi Thilanka Senadeera

Road feature detection from remotely sensed images is crucial for maintaining an up-to-date and reliable road network, essential for transportation, emergency planning, and navigation. While convolutional neural networks have shown promise in automating this process, existing methods often trade off accuracy for complexity. This study aims to develop an accurate road extraction method without sacrificing computational efficiency. We propose a semantic segmentation neural network combining transfer learning and U-net architecture with minimal complexity. Post-processing techniques are employed to enhance output quality. Our method achieves an F1 score of 0.83 and 95.57% accuracy, outperforming other models on the Massachusetts dataset. This approach demonstrates superior performance and reduced network complexity compared to existing methods.

During the recent years, it was occurred a large number of natural disasters throughout the world. One of the major reasons behind this is the disturbance in the equilibrium of the natural ecosystems. Deforestation can be identified as a main anthropogenic activity causing destruction of the natural environment. So, the studies on the deforestation have become a timely topic. Sri Lanka is one of the few remaining countries in the world with extensive natural forest cover, but much of the existing forests have been destroyed, mainly by shifting cultivation, logging and by the growth of number of people involved in agriculture. Therefore, the magnitude of deforestation and its consequences on natural processes need to be investigated. So, an attempt was taken in this research to understand the changes in forest cover throughout the study area using Microwave Remote Sensing techniques. The study area was selected in the Wilpattu National Park in Sri Lanka, because the area is highly deforested due to the human activities.