Computer Vision Pose Estimation for Hinged Vehicle Implement Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hinged vehicles, such as tractor trailers and construction tools, face challenges in accurately positioning implements due to uneven and sloped surfaces, as the positional independence between the vehicle and implement complicates precise alignment, often requiring expensive positioning devices like GNSS and IMU.
Innovation Solution
A pose estimating system using computer imaging, installed on the vehicle, captures images of the implement to calculate its orientation and position, allowing the vehicle to adjust its position to compensate for ground slope and undulations, thereby accurately positioning the implement without the need for expensive positioning devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive positioning devices like GNSS and IMU are used, then positioning accuracy is improved, but device cost and complexity increase
Solution Approach 1:
The patent uses image capture devices to create visual copies of the implement and its features, then processes these image copies to determine position and orientation. This replaces expensive physical positioning devices with optical copying and computational analysis, achieving high positioning accuracy without GNSS or IMU hardware on the implement.
Solution Approach 2:
The patent replaces mechanical positioning systems (GNSS receivers, IMU sensors) with an optical-computational system. Instead of using mechanical sensors to directly measure position and orientation, the system uses image capture devices to optically record the implement, then uses computer vision algorithms to calculate position and orientation from the images.
2Device complexity
If image capture devices and computer vision processing are used, then device complexity is reduced, but measurement precision may worsen
Solution Approach 1:
The patent segments the implement into identifiable features (edges, corners, markers) that can be independently detected in images. By breaking down the complex task of measuring entire implement position into smaller feature-point detections, the system achieves high precision using simple image processing of individual features rather than complex holistic measurement.
Solution Approach 2:
The patent changes the parameter space from direct physical measurements (requiring expensive sensors) to image domain parameters (pixel coordinates, edge orientations). By transforming the measurement problem into the image domain and using mathematical relationships between image features and real-world geometry, the system achieves high precision with low-cost imaging hardware.
Data Source
AI summary
Embodiments describe a method for positioning a hinged vehicle including a primary part and a secondary part coupled to the primary part at a project site. The method includes receiving, from an image capturing device, digital image data representing one or more features of the secondary part; performing image analysis on the digital image data to identify positions of the one or more features of the secondary part; identifying an angle of at least a portion of the secondary part; calculating a current position of the secondary part based on the angle; calculating a positional difference between a correct position at the project site for the secondary part and a current position of the secondary part at the project site; and initiating a change in a position of the primary part to compensate for the positional difference and to position the secondary part on the correct position.