Auto-Panning Camera View Using Image-Based Trailer Angle Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual panning systems in camera monitoring systems for commercial vehicles are inadequate for trailer reversing maneuvers due to inaccurate trailer angle estimates and insufficient granularity, leading to incomplete views for operators.
Innovation Solution
An automatic panning method that receives video feeds from cameras, identifies and tracks objects like wheels and lines, generates trailer angle measurements, and selects a single accurate angle using weighted scores, ensuring the trailer end remains in view and centered, utilizing image analysis and Kalman filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual panning systems are used to adjust camera views during trailer reversing, then the operator can adjust the view, but the system requires frequent stopping and provides insufficient granularity leading to incomplete views
Solution Approach 1:
The system automatically pans the camera view by detecting trailer angle from image data and calculating required camera adjustments, eliminating the need for manual operator intervention. The camera system self-adjusts to maintain optimal viewing of the trailer end during reversing operations.
Solution Approach 2:
The patent replaces manual mechanical panning with an automated image-based system that uses computer vision to detect trailer angle and automatically controls camera pan/tilt movements, substituting mechanical manual adjustment with automated digital control.
2Extent of automation
If automatic panning systems use traditional kinematic models with trailer angle estimates, then automation is achieved, but the trailer angle estimates are inaccurate leading to poor panning performance
Solution Approach 1:
The patent replaces traditional kinematic modeling with direct image-based geometric analysis. Instead of using vehicle sensor data and mathematical models to estimate trailer angle, the system directly detects the trailer angle from camera images by identifying trailer features and calculating their geometric relationships.
Solution Approach 2:
The patent introduces image-based trailer detection as an intermediary measurement method. Rather than directly measuring trailer angle through complex kinematic models, the system uses visual features in the camera image as an intermediate step to accurately determine the trailer angle.
3Quantity of substance
If multiple trailer angle measurements are generated from different object tracks, then more data is available, but selecting the accurate angle becomes complex requiring down-selection logic
Solution Approach 1:
The patent segments the trailer angle measurement problem into multiple independent object tracking tasks (wheels, lines, edges). Each object type provides its own angle measurement through separate detection pipelines, allowing the system to process different feature types independently and then select the most reliable measurement.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A method for automatically panning a view for a commercial vehicle includes receiving a video feed from at least one camera defining a field of view. A plurality of objects in the video feed, includes at least one wheel and at least one line, are identified. A path of each object in the plurality of objects is tracked through an image plane of the video feed. A trailer angle corresponding to each path is identified, the identified trailer angle is associated with the corresponding path, thereby generating a plurality of trailer angle measurements. The plurality of paths are down selected to a single path and identifying a single trailer angle measurement corresponding to the single path.