Onboard Parking Space Detection Using Visual Object Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting free parking spaces, whether using exteroceptive or proprioceptive sensors, face challenges such as high installation and maintenance costs, limited communication capabilities, and inefficiencies in detecting available spaces, particularly in real-time without external system reliance.
Innovation Solution
A method and device on a motor vehicle that captures images of its surroundings, utilizing a learning phase to store characteristics of known objects, enabling onboard image processing to detect and estimate the length of free parking spaces without external communication, using a camera to identify and correlate dimensions of objects like traffic signs, road markings, and vehicles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exteroceptive sensors (video cameras, ultrasonic sensors, magnetic sensors) are installed in the vicinity of parking spaces to detect free or occupied spaces, then detection capability is improved, but installation and maintenance costs increase and the parking space becomes unavailable during maintenance
Solution Approach 1:
The patent uses a video camera to capture optical images of the parking area and processes these images to detect free and occupied parking spaces. This replaces expensive exteroceptive sensors with a simpler imaging system that copies visual information, reducing hardware costs and maintenance requirements while maintaining detection capability.
Solution Approach 2:
The patent replaces physical sensors installed in parking spaces with an optical imaging system mounted on a mobile platform. This substitutes mechanical sensor installation with optical detection, eliminating the need for infrastructure installation and maintenance in parking spaces themselves.
2Speed
If exteroceptive sensors are used to detect parking space availability, then real-time detection is improved, but communication infrastructure requirements increase
Solution Approach 1:
The system performs self-service by capturing, processing, and analyzing images locally using an onboard processor. The mobile platform independently detects parking space availability without needing external communication infrastructure, eliminating dependencies on network systems while maintaining real-time detection capability.
3Loss of information
If proprioceptive sensors are used to evaluate probability that a vehicle will soon leave, then prediction capability is improved, but detection accuracy for already free spaces deteriorates
Solution Approach 1:
Instead of predicting future events (vehicle departure), the patent inverts the approach by directly detecting the current state (free parking spaces) through image analysis. The system processes visual data to identify occupied and free spaces in real-time, providing accurate detection of current availability rather than predicting future changes.
Data Source
Figure 1~3
Figure 4
AI summary
The invention relates to a method for detecting an available parking space from a motor vehicle being driven on a roadway and equipped with an image-capturing apparatus, said method comprising a learning phase in which characteristics relating to known objects are stored locally in a database (DB) of the vehicle, said method comprising: a step of acquiring an image of the roadway and its immediate environment; and a step of image processing, in which step: at least one known object, the characteristics of which have been learnt and stored beforehand, is detected (1.2); a parking area containing an available parking space is detected (1.3); and the length of the space is estimated (1.4) from a correlation between the characteristics, extracted from the database, of the known objects detected in the image.