Detecting Humans in Search and Rescue Operations Based on Ensemble Learning

Aim:

          To develop deep learning models to detect and track humans in aerial images

Abstract:

       Detecting humans in aerial images remains a tedious task for the application based on Search And Rescue operation (SAR). The prime goal of SAR is to detect and assist people who were met accident in mountain or other hazardous environment. For detecting people in SAR application, aerial image of mountain landscapes is utilized. The major challenges in detecting human in aerial images are pose and scale variations of humans, low visibility, camouflaged environment, adverse weather conditions, motion blur, and high-resolution aerial images. Due to imaging from high altitudes, only 0.1 to 0.2 percentage of the image represents humans. To solve the problem of low coverage of the object of interest in high-resolution aerial images, we propose to implement a deep learning-based object detection model. In this model, we propose a novel method for the detection of humans in aerial images based on Deep learning architecture.

Synopsis:

       Object detection is one of the most researched areas in computer vision. It is the process of determining where exactly the object is in the scene or image and what object has been detected. Object detection in aerial images depends on several factors such as low visibility due to varying altitudes, the object-of-interest, variations in pose and scale, camouflaged environment with rocks and trees, and high-resolution aerial image.

Proposed System:

          In the proposed system, we have used yolo model v5 in this project. It gives better accuracy compared to other and we have able to predict it better in aerial images. YOLO V5 is fast and accurate compared to other models.

Advantage:

       YOLOv5 is a modern object detection algorithm, that has been written in a PyTorch, Besides this, it’s having, fast speed, high accuracy, easy to install and use

YOLO V5:

     YOLOv5 is a model in the You Only Look Once (YOLO) family of computer vision models. YOLOv5 is commonly used for detecting objects. YOLOv5 comes in four main versions: small , medium, large, and extra large, each offering progressively higher accuracy rates. YOLO v5 uses a new method for generating the anchor boxes, called “dynamic anchor boxes.” It involves using a clustering algorithm to group the ground truth bounding boxes into clusters and then using the centroids of the clusters as the anchor boxes.