Monocular Airborne Object Detection via Motion Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unmanned aerial vehicles (UAVs) face challenges in detecting airborne obstacles due to power and weight constraints, making traditional sensing technologies like stereoscopic image processing and single sensor technologies such as Radar and Lidar unsuitable for lightweight designs.
Innovation Solution
A monocular airborne object detection system using an imaging camera, inertial measurement unit, and a computer system that generates a motion-compensated background image sequence by calculating transformations between image frames, allowing for the differentiation of moving objects from static ones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stereoscopic image processing is used to detect airborne objects, then depth resolution is improved, but device weight and complexity increase due to requiring duplicate image sensors
Solution Approach 1:
The patent segments the depth measurement function into temporal domains by capturing multiple images at different time points, replacing the spatial segmentation required by stereoscopic sensors. This allows a single sensor to achieve depth information through sequential temporal sampling rather than simultaneous spatial sampling.
Solution Approach 2:
The patent transitions from spatial dimension (using multiple sensors at different positions) to temporal dimension (using a single sensor at multiple time points) to achieve the same depth measurement capability. This dimensional transformation eliminates the need for duplicate sensors while preserving depth resolution.
2Measurement precision
If stereoscopic image processing is used to detect airborne objects, then depth resolution is improved, but device complexity increases due to requiring duplicate image sensors
Solution Approach 1:
The patent segments the depth measurement function into temporal domains by capturing multiple images at different time points, replacing the spatial segmentation required by stereoscopic sensors. This allows a single sensor to achieve depth information through sequential temporal sampling rather than simultaneous spatial sampling.
Solution Approach 2:
The patent transitions from spatial dimension (using multiple sensors at different positions) to temporal dimension (using a single sensor at multiple time points) to achieve the same depth measurement capability. This dimensional transformation eliminates the need for duplicate sensors while preserving depth resolution.
3Reliability
If single sensor technologies such as Radar and Lidar are used, then detection capability is improved, but device weight exceeds available dimensions for lightweight airborne UAV designs
Solution Approach 1:
The patent replaces active sensing technologies (Radar, Lidar) that emit electromagnetic waves or light with passive optical imaging that captures reflected light. This substitution eliminates the need for heavy power sources and active sensors, using instead a lightweight camera system that processes visual information through computational algorithms.
Solution Approach 2:
The patent creates temporal copies of the same scene at different time points using a single camera, replacing the need for physical copies (multiple sensors) or active illumination systems. The computational processing of these temporal copies provides detection capability without the weight penalty of traditional active sensors.
4Reliability
If single sensor technologies such as Radar and Lidar are used, then detection capability is improved, but power consumption increases due to power sources required to power such equipment
Solution Approach 1:
The patent replaces active sensing technologies (Radar, Lidar) that emit electromagnetic waves or light with passive optical imaging that captures reflected light. This substitution eliminates the need for heavy power sources and active sensors, using instead a lightweight camera system that processes visual information through computational algorithms.
Solution Approach 2:
The patent creates temporal copies of the same scene at different time points using a single camera, replacing the need for physical copies (multiple sensors) or active illumination systems. The computational processing of these temporal copies provides detection capability without the weight penalty of traditional active sensors.
Data Source
AI summary
Systems and methods for airborne object detection using monocular sensors are provided. In one embodiment, a system for detecting moving objects from a mobile vehicle comprises: an imaging camera; a navigation unit including at least an inertial measurement unit; and a computer system coupled to the image camera and the navigation unit. The computer system executes an airborne object detection process algorithm, and calculates a transformation between two or more image frames captured by the imaging camera using navigation information associated with each of the two or more image frames to generate a motion compensated background image sequence. The computer system detects moving objects from the motion compensated background image sequence.