Developed a warehouse management drone system where a drone autonomously takes off from a set location, navigates toward shelves/pallets, scans positions, collects data, and returns to its base. Built using PX4, ROS, Python, and Gazebo simulation, with custom control algorithms for precise movement and navigation. While the scanning component was not implemented in this phase, the system is designed to integrate RGB camera scanning for full warehouse automation.

Developed an obstacle avoidance system for drones using ArduPilot firmware and custom control algorithms. A stereo camera was integrated for real-time depth estimation, enabling the drone to detect obstacles and plan safe navigation paths. Implemented path planning techniques to ensure smooth, autonomous flight in static/dynamic environments.

Developed a fully functional e-commerce website with a user-friendly interface and secure backend. Implemented features like product listings, shopping cart, order management, and payment integration. Built with React, Node.js, and scalable backend APIs, ensuring smooth performance, secure transactions, and reliable hosting for high availability.

Implemented algorithms to process aerial drone imagery into highly accurate 3D models and orthomosaic maps. Using photogrammetry techniques such as Structure-from-Motion (SfM) and Multi-View Stereo (MVS), I reconstructed dense point clouds and textured 3D meshes that capture the vertical details of terrain and infrastructure. In parallel, I designed an image stitching and orthorectification pipeline that combined multiple overlapping drone images into a seamless, geo-referenced orthomosaic map with uniform scale and true spatial accuracy.

Developed an intelligent vision pipeline for real-time object detection and tracking from imagery data. Implemented deep learning models such as YOLO, SSD, and custom models to detect vehicles, humans, and infrastructure elements, and integrated tracking algorithms to follow moving objects across frames.

The system enables to not only identify objects of interest but also continuously monitor their movement, making it highly applicable for security surveillance, traffic monitoring, search-and-rescue for autonomous vehicles, and inspection tasks.

Designed and implemented a system for real-time video streaming over the cloud. The project focused on capturing live camera feeds and transmitting them securely using protocols like RTSP/RTMP to a cloud server. The video stream was then processed, optimized, and made accessible through a web dashboard, allowing remote users to monitor drone operations in real time.

This solution enables remote surveillance, live inspections, and collaborative monitoring, reducing the need for on-site presence and ensuring faster decision-making.