Forward-Looking Camera Terrain Matching for Missile Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low-flying missiles face navigation challenges due to disruptions in satellite navigation systems and errors in inertial navigation systems, especially on long flight routes, and existing map-matching navigation methods are impractical with forward-looking cameras.
Innovation Solution
A method that determines the current position and attitude of a missile by comparing 3D reconstructions from a forward-looking camera's image sequence with 2D reference material, using line segments extracted via standard methods and synchronized with inertial measurement data, to correct navigation errors without GPS support or downward-facing cameras.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If satellite navigation systems (GPS) are used for missile navigation, then navigation accuracy is improved, but the system is easy to disrupt particularly on low-flying missiles
Solution Approach 1:
The patent introduces an intermediary system consisting of a forward-looking camera and terrain database that acts as a mediator between the missile and the ground terrain. The camera captures terrain images that are compared with pre-stored terrain database images, providing navigation correction without direct satellite dependency. This intermediary terrain-matching mechanism resolves the contradiction by providing reliable navigation through indirect terrain observation rather than direct satellite signal reception.
2Reliability
If inertial navigation systems are used for missile navigation, then the system is reliable and works without external support, but navigation errors accumulate on long flight routes
Solution Approach 1:
The patent implements a feedback mechanism where terrain images captured by the forward-looking camera are continuously compared with corresponding images from the terrain database. The position and orientation differences between matched terrain features provide feedback signals that correct the drift errors accumulating in the inertial navigation system. This feedback loop maintains navigation accuracy over long flight routes while preserving the reliability of independent inertial operation.
3Measurement precision
If map-matching navigation with downward-facing camera is used, then navigation correction is possible, but the camera must be directed vertically downwards which conflicts with forward-looking seeker cameras
Solution Approach 1:
The patent inverts the conventional map-matching approach by using a forward-looking camera instead of a downward-facing camera. Rather than projecting the missile's position down onto the terrain below, the system projects the forward-looking camera's terrain observations onto the terrain database. This inversion allows the use of forward-looking seeker cameras for both guidance and navigation correction functions, resolving the conflict between camera orientation requirements.
4Adaptability or versatility
If forward-looking camera is used for terrain observation, then the camera can be integrated with seeker heads, but existing map-matching methods cannot be applied due to different image characteristics
Solution Approach 1:
The patent applies parameter changes by transforming the forward-looking camera images into a coordinate system and projection perspective that matches the terrain database. The system performs geometric transformations including perspective projection, coordinate rotation, and scale adjustment to make forward-looking images comparable with the stored terrain references. This parameter transformation enables image comparison despite the different viewing angles and characteristics, allowing successful integration of forward-looking cameras with terrain-matching navigation.
Data Source
AI summary
Method for supporting the autonomous navigation of a flying object having a forward-looking camera and an INS system by comparing data from the current scene recorded by the camera with reference images available to the flying object, wherein a 3D structure of the scene is obtained from the recorded image sequence of the overflown scene by tracking and 3D reconstruction of significant line segments, which, after projection into the reference perspective, is compared with available 2D reference material in order to determine the current position and orientation of the flying object for the purpose of supporting its navigation.