Trailer Camera Position Shift for Low-Speed Object Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems fail to accurately detect objects, such as people lying on the ground, when vehicles are stationary or moving at very low speeds, due to limitations in spatial mapping and object classification using a single camera.

Innovation Solution

A method utilizing a trailer-mounted camera and an active actuator system to adjust its position, combined with odometry data including articulation angles, enables accurate object detection and depth information determination through Structure-from-Motion (SfM) by capturing images from different viewpoints.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple sensors are used to detect objects in the vehicle environment, then the reliability of object detection is improved, but the device complexity increases

Engineering Contradiction:
Improveobject detection reliabilityVSAvoidsensor system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent combines data from multiple sensors (ultrasonic, radar, camera) into a unified sensor data set, processing them together through a single neural network model rather than treating them as separate systems. This merging approach maintains the reliability benefits of multiple sensors while reducing overall system complexity.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The neural network model is designed to process multiple types of sensor data (ultrasonic, radar, camera) through a unified architecture, making the system multi-functional. The same processing pipeline handles different sensor types, reducing the need for separate processing systems for each sensor type.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Measurement precision

If sensor data is processed in high resolution to improve object detection accuracy, then the measurement precision is improved, but the loss of time increases due to higher computational requirements

Engineering Contradiction:
Improveobject detection precisionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system applies resolution reduction selectively - not all sensor data is processed at full resolution. The neural network processes a down-sampled version of the sensor data set, using partial action (reduced resolution) to achieve acceptable detection precision while significantly reducing processing time and computational load.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent changes the resolution parameter of the sensor data before processing. By down-sampling the sensor data set to a lower resolution, the system reduces the computational complexity of neural network processing while maintaining sufficient detection accuracy for safety-critical applications.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If the neural network model is trained with high computational resources to improve detection accuracy, then the measurement precision is improved, but the loss of energy increases during training

Engineering Contradiction:
Improveobject detection precisionVSAvoidtraining energy consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system uses partial action in training by processing down-sampled sensor data during the training phase. The neural network is trained on reduced-resolution data sets, which requires significantly less computational power and energy while still achieving sufficient detection accuracy for the application.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent applies parameter changes by modifying the resolution of training data. The neural network model is trained on down-sampled sensor data rather than full-resolution data, reducing the energy consumption of training while maintaining adequate detection precision for safety applications.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4118622B1Method for determining object information relating to an object in a vehicle environment, control unit and vehicle
Publication Date: 2026.04.29 ZF CV SYST GLOBAL GMBH
  • EP4118622B1 patent drawingFigure 1~1a
  • EP4118622B1 patent drawingFigure 2a~2b
  • EP4118622B1 patent drawingFigure 3

AI summary

The invention relates to a method for determining object information relating to an object in an environment of a multi-part vehicle, formed by at least one towing vehicle and at least one trailer, as well as a control unit and vehicle for same, wherein at least one trailer camera is arranged at least on the trailer, comprising at least the following steps: capturing the environment using a trailer camera from a first position and creating a first image consisting of first pixels; changing the position of the trailer camera; capturing the environment using the trailer camera from a second position and creating a second image consisting of second pixels; and determining object information relating to an object in the captured environment.