Problem: Growing Wildfire has come a global concern and there is an immediate need of developing image processing based efficient technology for its early detection. Traditional models can only classify Fire and No fire without the consideration of smoke that too with a low accuracy rate.
Project: This project uses ResNet50 (residual network) to classify a image into fire,neutral and smoke. Detection of smoke is signification as it can create alert for fires and early detection can help in reducing the damage of life and property
Results Obtained:
Dataset link - https://drive.google.com/drive/folders/1-3jebe6EG8r5rBJ_YV3daebqdF9rKKmH?usp=drive_link
Website Link: https://sites.google.com/vt.edu/wildfire