Autonomous Drone Image Navigation Without GPS or Multi-Sensor Fusion

The patent extracts and removes the GPS dependency and multiple sensor requirements from the navigation system. Instead of using complex multi-sensor fusion, the invention relies on a single camera to capture images and determine positional information through image processing alone, thereby simplifying the device while maintaining navigation capability

The patent replaces the mechanical/sensor-based navigation system with an optical/image-based system. By substituting physical sensors and GPS hardware with image capture and computational image analysis, the system achieves navigation functionality with reduced hardware complexity and power consumption

The patent removes power-intensive components (GPS receiver, multiple sensors) from the system and replaces them with a more energy-efficient single-camera solution that achieves navigation through image processing, thereby reducing overall power consumption while maintaining navigation accuracy

The patent substitutes power-consuming hardware sensors with an optical/image-based navigation approach, where a camera captures environmental features and computational algorithms determine position and orientation, resulting in lower energy consumption compared to traditional sensor fusion systems

The patent performs preliminary stabilization actions during the takeoff phase by hovering and capturing images to establish initial positional reference. This preliminary action allows the drone to achieve accurate stabilization quickly, after which it can proceed with the flight plan without prolonged hovering, thereby preserving flight time

The patent applies partial stabilization hovering - just enough image capture during takeoff to establish initial position and achieve required stabilization accuracy, rather than continuous hovering throughout the flight. This partial action approach achieves the necessary stabilization while minimizing time loss and preserving overall flight duration