This is my undergraduate thesis project
Remote sensing is a vital tool for academics and researchers in agriculture, urban planning, public health, and many more. A trained eye can extract helpful information where any ordinary person will see an aerial picture. However, acquiring images directly from a satellite is not always possible. Again, many do not have access to such data due to their limitations, such as time or funds. Unmanned Aerial vehicles or u A v has proved a reliable aerial photography tool. Most aerial mapping systems utilizing UAV or remote-controlled (RC) multi-rotor aircraft are costly and require a specialist team to operate them. Again cheaper, easy to fly, and widely available remote-controlled (RC) multi-rotor aircraft are intended for children and hobbyists. They do not offer the quality or the functionalities required for a scientific project. In such specific cases, where aerial imaging is required on low budget, only solution is to build a custom multi-rotor aircraft, tuned precisely for scientific purposes. Unfortunately, it still requires a team who are skilled builders and experienced RC pilots. o u r project aims to bridge the gap that exists. o u r project goal is to provide a platform for remote sensing based on UAVs for developing countries such as Bangladesh where the budget is not abundant. We have devised a cost-efficient way to build semi-automated UAVs that are easy to operate for a novice RC pilot, and have the necessary functionality a researcher will likely need to complete his experimentation or data collection. Our design utilizes local components to build the main frame for a quadcopter to keep the cost down, ArduPilot for high precision control to ensure the safety of surrounding people, and a software interface that delivers cutting-edge performance almost free of cost. For control and communication link with our UAV, we used a 2.54 GHz band, which is free of cost and produces satisfactory results for operation within line of sight. We have used open CV, an open-sourced, community-maintained computer vision library, to create a primary image processing platform. Altogether we have been able to create a cost-effective and efficient system that can be easily adapted to any form of aerial imaging where both time and resources are limited.