• Amir Azizi

Building damage assessment after natural disasters is a critical task in the recovery process. In this paper, we present a machine learning-based approach for multilabel classifying damaged buildings using UAV imagery and image processing techniques. Our method utilizes the RescueNet image dataset collected with DJI Mavic Pro quadcopters after Hurricane Michael. The dataset contains 4494 images divided into training, validation, and test sets. We employ a pre-trained deep neural network based on the CLIP (Contrastive Language-Image Pre-Training) method and fine-tune it on the RescueNet dataset to achieve accurate building damage assessments. Our experimental results, compared with state-of-the-art methods such as YOLOV8, EfficientNet, and MobileNetV2, indicate that our proposed method achieves the best performance in terms of accuracy and speed, demonstrating its superiority for object detection tasks. The final accuracy of our method is 92%, which demonstrates its effectiveness in real-world scenarios. results show that the proposed approach provides a fast and reliable solution for building damage assessment and has the potential to be widely applied in disaster management.

building damage assessment-UAV imagery-Image processing-Multi label classification-Machine learning

"A CLIP-Based Method for Efficient Building Damage Assessment using UAV Images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 7, page no.a678-a686, July-2023, Available :http://www.jetir.org/papers/JETIR2307082.pdf

"A CLIP-Based Method for Efficient Building Damage Assessment using UAV Images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 7, page no. ppa678-a686, July-2023, Available at : http://www.jetir.org/papers/JETIR2307082.pdf