Parking Space Recognition With Multi-Angle Reference Image Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing parking assistance systems face accuracy issues in recognizing target parking spaces due to images lacking features suitable for recognition processing, leading to decreased recognition accuracy.
Innovation Solution
The system stores multiple reference images of a target parking space captured from different viewpoints and angles, using these images to verify the current image and autonomously control the vehicle's travel to the target space based on recognized positional information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single overhead image is generated and used for parking space recognition, then the system complexity is reduced, but the recognition accuracy deteriorates when the image lacks suitable features
Solution Approach 1:
The patent segments the reference image into multiple candidate parking space regions, each with distinct features. Instead of treating the entire image as one unit, the system divides it into separable candidate areas that can be independently verified, allowing the system to handle complex recognition scenarios without requiring overly complex processing of the whole image at once.
Solution Approach 2:
The patent introduces a verification dimension by comparing candidate parking spaces against stored template images. This adds a comparative verification layer that enhances recognition accuracy without significantly increasing the base system complexity, as it uses existing template matching techniques in a new verification context.
2Measurement precision
If multiple reference images are stored and used for verification, then the recognition accuracy is improved, but the data storage requirements and processing complexity increase
Solution Approach 1:
The patent performs preliminary segmentation of reference images into candidate parking space regions during the image capture phase. By pre-identifying and storing only the relevant candidate areas rather than entire images, the system reduces data storage requirements while maintaining the ability to perform accurate verification when needed.
Solution Approach 2:
The patent extracts and stores only the essential candidate parking space regions from reference images, separating the relevant recognition features from unnecessary image data. This extraction process reduces the quantity of stored data while preserving the critical information needed for accurate parking space verification.
3Productivity
If template matching is used for parking space verification, then the recognition speed is improved, but the accuracy deteriorates when the current image conditions differ from template conditions
Solution Approach 1:
The patent dynamically adjusts the verification process by first performing rapid template matching to identify potential parking spaces, then applying feature-based verification to confirm accuracy. This dynamic two-stage approach maintains high recognition speed through efficient template matching while ensuring accuracy through adaptive feature verification when conditions warrant it.
Solution Approach 2:
The patent implements feedback by using template matching results to guide subsequent feature verification. The template matching provides initial candidates that are then verified against actual image features, creating a feedback loop where the results of one verification method inform and refine the other, thereby maintaining both speed and accuracy.
Data Source
AI summary
A plurality of different reference images for specifying a target parking space is preliminarily stored. The different reference images are generated by using an imaging device that captures an image around a subject vehicle. After that, when the subject vehicle is parked in the target parking space, the current image captured around the subject vehicle is verified with at least one reference image of the plurality of different reference images to recognize the target parking space, and the travel operation of the subject vehicle to the target parking space is autonomously controlled based on positional information of the recognized target parking space.