Change Detection and Budget Estimation of Catastrophic Events Based on Image Processing

Each year, there are hundreds of natural disasters worldwide. For disaster relief and recovery operations, pinpointing high-impact areas and accurately analyzing the post-disaster landscape are essential. It would make it possible to evaluate the damage and pinpoint the best paths for quick assistance in crisis situations. Currently, volunteer efforts around the world employ satellite imagery to carry out the necessary mapping following natural catastrophes that result in humanitarian situations. However, the process takes time and varies across different projects because the volunteers tend to be inexperienced and adhere to varied procedures. It may be possible to spot places that have altered by comparing the pixel values of satellite photos taken before and after the tragedy. Compared to human efforts, the Change Detection technique can compare pixel values more successfully. The processes that result in changes in land use or land cover are identified by looking at changes in a region's spectral features over time. Change detection examines the variation in photographs taken over a predetermined time period and in relation to a predetermined location. The occurrence and extent of the harm are determined by this change in the image. A change detection-based model is used in this project to recognize disaster occurrence, forecast the cost to repair the damage, and estimate the percent likelihood of occurrence.