MooketsiKosa/SentinalSA

JavaFX disaster response tool that converts aerial imagery into a weighted graph and finds the safest rescue route using Dijkstra's algorithm. Built for South African emergency response scenarios (KZN floods, informal settlement fires). # SentinalSA JavaFX disaster response tool that converts aerial imagery into a weighted graph and finds the safest rescue route using Dijkstra's algorithm. Built for South African emergency response scenarios (KZN floods, informal settlement fires). # SentinelPath SA > **Real-time disaster response pathfinding for South African emergency teams.** > Converts aerial imagery into a risk-weighted graph and calculates the safest rescue route — even when standard maps fail. ## The Problem South Africa regularly faces localised disasters the 2022 KZN floods, fires in informal settlements where road infrastructure is destroyed and conventional mapping tools become unreliable. Emergency responders need a way to navigate unpredictable, rapidly changing terrain using only what they can see from the air. **SentinelPath SA** solves this by treating any aerial `.jpg` or `.png` as a live navigable map, classifying terrain by colour, building a weighted graph on-the-fly, and computing the safest path for rescue vehicles in real time. ## How It Works ### 1. Image → Graph Abstraction | Concept | Implementation | | **Nodes (V)** | Image segmented into a 50 × 50 grid; each cell = one node | | **Edges (E)** | Each node connects to its 8 neighbours (horizontal, vertical, diagonal) | | **Weights (W)** | Edge cost = terrain classification of the target node | ### 2. Terrain Classification (Task 1) - A **Hash Table** stores reference RGB values for known terrain types. - Each grid node's colour is sampled and matched to the nearest reference category. - Classification drives traversal cost assignment: | Terrain | Weight | | Road | 1 | | Vegetation / Grass | 5 | | Debris / Unknown | 50 | | Water | 1,000 | ### 3. Pathfinding — Dijkstra's Algorithm (Task 2) - Implemented with a **manual Adaptable Priority Queue** (min-heap). - Complexity: **O(E log V)** handles the full 50×50 grid interactively. - The safest path updates instantly as users modify the map. ## Features - **Interactive Map C …