The Micro-Doppler Based Target Classification project focuses on distinguishing between different Small Aerial Targets like Drones, RC Planes, Birds, etc by analyzing micro-Doppler signatures captured from radar sensors.
This technology is essential for improving situational awareness in scenarios such as surveillance, search and rescue missions, and safeguarding critical infrastructure.
The project encompasses:
Signal Processing: Obtained dataset from IEEE Dataport, which is around 4000 spectrogram images extracted from radar data. This data is normalized, and augmented. Feature extraction is used to to extract meaningful features from the spectrograms, to enhance model's performance.
Machine Learning: Implementing Python-based models to classify objects based on the extracted features.
Web Application: Creating an intuitive interface for users to interact with the classification system.