Camera-Based Vehicle Navigation With Trajectory-Guided Feature Tracking

Solution Overview

Problem

Conventional vehicle vision systems face inefficiencies in tracking features when vehicles are moving on non-straight trajectories or at non-constant speeds, leading to inaccurate motion estimates and increased computational resource requirements.

Innovation Solution

The vehicular vision system employs an attentiveness algorithm that adjusts the region of interest based on the predicted vehicle trajectory, prioritizing feature tracking in areas relevant to the future path, thereby optimizing computational resources and reducing latency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the system tracks all detected features in subsequent frames, then measurement precision is improved, but computational resource requirements increase and processing latency increases

Engineering Contradiction:

Improvemotion estimation accuracyVSAvoidprocessing efficiency

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the feature tracking task by dividing all detected features into two subsets: tracked features and untracked features. This segmentation allows the system to process only the most relevant features (those likely to be in the vehicle's future path) while ignoring others, thereby reducing computational load and processing latency while maintaining sufficient motion estimation accuracy for safe autonomous operation.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by differentiating the processing quality applied to different features. Tracked features receive full attention and precise tracking, while untracked features receive minimal or no tracking resources. This differentiated approach optimizes computational resource allocation by concentrating processing power on features that are most critical for the vehicle's predicted trajectory.

Inventive Principle:

Principle #3Local quality

2Measurement precision

If the system tracks all detected features, then measurement precision is improved, but loss of time increases due to increased processing latency

Engineering Contradiction:

Improvemotion estimation accuracyVSAvoidprocessing latency

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the feature tracking task by dividing all detected features into two subsets: tracked features and untracked features. This segmentation allows the system to process only the most relevant features (those likely to be in the vehicle's future path) while ignoring others, thereby reducing computational load and processing latency while maintaining sufficient motion estimation accuracy for safe autonomous operation.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary action by predicting the vehicle's future trajectory before tracking features. This prediction is used to pre-identify which features are likely to be relevant, allowing the system to select the subset of tracked features in advance. This preliminary selection reduces the computational burden during the actual tracking phase, thereby reducing processing latency.

Inventive Principle:

Principle #10Preliminary action

3Measurement precision

If the system processes all features, then measurement precision is improved, but device complexity increases due to higher computational resource requirements

Engineering Contradiction:

Improvemotion estimation accuracyVSAvoidcomputational resource requirements

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the feature tracking task by dividing all detected features into two subsets: tracked features and untracked features. This segmentation allows the system to process only the most relevant features (those likely to be in the vehicle's future path) while ignoring others, thereby reducing computational load and processing latency while maintaining sufficient motion estimation accuracy for safe autonomous operation.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by differentiating the processing quality applied to different features. Tracked features receive full attention and precise tracking, while untracked features receive minimal or no tracking resources. This differentiated approach optimizes computational resource allocation by concentrating processing power on features that are most critical for the vehicle's predicted trajectory.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20240109545A1Vehicular control system utilizing cameras for vision-based navigation of the vehicle

Publication Date: 2024.04.04 MAGNA ELECTRONICS INC

AI summary

A vehicular driving assist system includes a camera disposed at a vehicle and capturing frames of image data. An electronic control unit (ECU) includes an image processor for processing frames of image data captured by the camera. The system, responsive to processing at the ECU of frames of image data captured by the camera, detects a plurality of features, each feature representative of at least a portion of an object viewed by the camera. The system predicts a trajectory of the vehicle and selects a subset of the plurality of features based on the predicted trajectory of the vehicle. The system (i) tracks the subset of the plurality of features in subsequent frames of image data captured by the camera and (ii) does not track features outside the subset of the plurality of features in subsequent frames of image data captured by the camera.