Parking Assist Feature Point Updating Under Changing Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing parking assist systems face challenges in accurately identifying parking positions when feature points are not detected due to shade, dirt, or changes in environmental conditions, leading to incorrect feature point extraction and subsequent failure in automatic parking.
Innovation Solution
A parking assist apparatus equipped with a camera and controller that registers feature points, erases undetectable points after consecutive failures, and re-extracts new feature points to maintain accurate parking position calculation, ensuring the vehicle can properly park even with changes in environmental conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If feature points are extracted from captured images during vehicle movement, then automatic parking can be achieved using structure positions, but detection reliability deteriorates when shade or environmental changes occur
Solution Approach 1:
The system performs preliminary extraction of multiple candidate feature points from captured images before automatic parking execution. By pre-identifying and storing multiple potential feature points with their coordinates and validity flags, the system ensures that valid feature points are ready for use even when environmental conditions change during parking execution, thus maintaining detection reliability while enabling automatic parking.
Solution Approach 2:
The system dynamically changes the parameter of feature point selection by validating candidate points against current image data. When shade or environmental changes cause original feature points to become undetectable, the system changes which feature points are active by clearing validity flags on failed points and selecting alternative valid points from the pre-extracted candidates, thereby maintaining detection reliability under varying conditions.
2Measurement precision
If the same feature points are reused for parking calculation, then parking position accuracy is maintained, but adaptability deteriorates when environmental conditions change over time
Solution Approach 1:
The system implements feedback by detecting whether pre-extracted feature points are still visible and valid in current images during automatic parking. When a feature point becomes undetectable due to environmental changes like shade or dirt, the system receives feedback through the validity check, clears the point's validity flag, and selects alternative feature points, thus maintaining both calculation accuracy and adaptability to environmental changes.
Solution Approach 2:
The system makes the feature point set dynamic by allowing the composition of valid feature points to change based on current detection results. Instead of using a fixed set of feature points, the system dynamically updates which points are valid and usable for parking calculation, enabling adaptation to environmental changes while maintaining precision through continuous validation and replacement of invalid points.
3Reliability
If multiple candidate feature points are extracted and stored, then detection reliability improves under varying conditions, but device complexity increases
Solution Approach 1:
The system extracts and separates the complexity of feature point management by pre-extracting multiple candidate feature points with their coordinates and validity information, storing them as discrete data structures. This extraction of candidate points before execution allows the system to manage complexity through organized data storage rather than complex real-time processing, improving reliability while controlling device complexity through structured data management.
Data Source
AI summary
A parking assist apparatus including a controller configured to perform operations including image feature extraction, parking feature registration, identification of a parking position from the registered parking features, performance of parking assistance based on the identified parking position, and erasure of registered parking features based on non-detection of the parking features during the performance of parking assistance.