Tow Ball Detection Using Convolutional Feature Patches

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing trailer hitch assist systems face inaccuracies, high costs, and time consumption when manually measuring the position of the tow hitch, especially when the tow coupling is not mounted during vehicle manufacturing or can be moved manually or electronically.

Innovation Solution

A method for determining the position of a tow ball in an image using real-time images from a vehicle, where the images are processed to extract feature patches, and a convolution is performed with filters based on known tow hitches to identify the tow ball location.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual measurement is used to determine tow hitch position, then simplicity is maintained, but accuracy deteriorates and time consumption increases

Engineering Contradiction:
Improvetow hitch position accuracyVSAvoidtime consumption
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical measurement with an automated image processing system. A camera captures images of the tow hitch, and a controller processes these images to automatically determine the tow hitch position, eliminating the need for manual measurement while improving both accuracy and speed.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent uses visual copies (images) of the tow hitch instead of direct physical measurement. By capturing the tow hitch in an image and processing the image data, the system determines position without physical contact or manual intervention, thereby reducing time consumption while maintaining precision.

Inventive Principle:
Principle #26Copying

2Measurement precision

If manual measurement is used to determine tow hitch position, then device complexity is reduced, but manufacturing cost increases

Engineering Contradiction:
Improvetow hitch position accuracyVSAvoidmanufacturing cost
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The system enables the vehicle to self-determine its tow hitch position using onboard cameras and controllers. This self-service approach eliminates the need for expensive manual measurement services during manufacturing, reducing overall costs while maintaining high accuracy through automated image processing.

Inventive Principle:
Principle #25Self-service

3Measurement precision

If automated image processing is used to determine tow hitch position, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvetow hitch position accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the image processing task into distinct components: capturing images with a camera, extracting features from the images, and processing these features to determine position. This segmentation allows each component to be optimized independently and simplifies the overall system architecture while maintaining high precision.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4012656B1Method for real-time tow ball detection
Publication Date: 2025.05.07 AUMOVIO AUTONOMOUS MOBILITY US LLC
  • EP4012656B1 patent drawingFigure 1~2
  • EP4012656B1 patent drawingFigure 3
  • EP4012656B1 patent drawingFigure 4~6

AI summary

A method (100) for determining a position of a tow ball (26) in an image. The method (100) includes obtaining at least one real-time image (40) from a vehicle (20). The at least one real-time image (40) is processed with a controller (30) on the vehicle (20) to obtain a feature patch (f) describing at least one real-time image (40). A convolution is performed of the feature patch (f) and each filter (h) from a set of filters (H) with the filter (h) being based on data representative of known tow hitches fixed to the vehicle (20). A location of a tow ball (26) on the tow hitch (22) is identified in the at least one real-time image (40) is based on the convolution between the feature patch (f) and each filter (h) from the set of filters (H).