Autonomous UAV Image Capture With Visual Tracking and Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current unmanned aerial vehicle (UAV) systems for image and video capture require piloting expertise and are prone to crashes due to pilot error, as they necessitate direct control of pitch, roll, yaw, and power adjustments, which can be complex and error-prone.
Innovation Solution
An autonomous UAV system that uses visual inertial odometry, computer vision, and a network of phased array wireless transceivers for localization and navigation, allowing it to adjust its flight path and image capture settings dynamically to maintain a clear line of sight with a subject and avoid obstacles without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If direct control of pitch, roll, yaw, and power is used for UAV operation, then the pilot can manually adjust flight parameters, but the system becomes prone to crashes due to pilot error and requires extensive piloting expertise
Solution Approach 1:
The UAV system performs self-localization using visual inertial odometry and self-navigation using computer vision algorithms to detect and track subjects, enabling the system to autonomously adjust its flight path and capture settings without requiring manual piloting intervention
Solution Approach 2:
The patent replaces manual mechanical control (joystick, throttle) with automated computer vision-based control systems that use image processing and machine learning algorithms to automatically determine flight adjustments and subject tracking
2Reliability
If automated localization and navigation systems are implemented, then pilot error is reduced and crash resistance improves, but the device complexity increases due to additional sensors and processing requirements
Solution Approach 1:
The UAV's image capture devices serve dual purposes: they capture images for their primary function and simultaneously provide visual data for inertial odometry calculations and computer vision-based subject detection, eliminating the need for separate dedicated sensors
Solution Approach 2:
The patent combines multiple functions (image capture, localization, navigation, subject tracking) into a unified system where the same hardware components (cameras, processors) perform multiple tasks through software algorithms rather than requiring separate physical systems
3Manufacturing precision
If the UAV autonomously adjusts flight path and image capture settings to maintain clear line of sight with subject, then image quality improves, but the extent of automation increases requiring sophisticated algorithms
Solution Approach 1:
The system continuously captures images, processes them through computer vision algorithms to detect subject position and environmental features, and uses this feedback to automatically adjust flight path and camera settings in real-time to maintain optimal subject framing and image quality
Solution Approach 2:
The UAV performs preliminary localization and subject detection before executing flight maneuvers, using visual inertial odometry to pre-calculate optimal flight paths that will maintain clear line of sight with the subject while avoiding obstacles
Data Source
AI summary
Methods and systems are disclosed for an unmanned aerial vehicle (UAV) configured to autonomously navigate a physical environment while capturing images of the physical environment. In some embodiments, the motion of the UAV and a subject in the physical environment may be estimated based in part on images of the physical environment captured by the UAV. In response to estimating the motions, image capture by the UAV may be dynamically adjusted to satisfy a specified criterion related to a quality of the image capture.