Ultrasonic Parking Detection Using Echo Signal Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing parking space detection systems using ultrasonic sensors have a high erroneous detection rate, often identifying reflections from trees or walls as actual parking spaces, leading to inaccurate parking space identification.
Innovation Solution
The method employs environment sensors, such as radar, lidar, or ultrasonic sensors, to detect objects and classify them based on characteristic structures of echo signals, determining parking space quality by evaluating the spatial arrangement and properties of objects delimiting the space, thereby reducing incorrect detections and providing a multi-level quality measure for filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If ultrasonic sensors are used to detect parking spaces, then parking space detection capability is provided, but the false detection rate increases
Solution Approach 1:
The patent changes the evaluation parameters from simple distance measurement to analyzing characteristic structures of echo signals (hyperbolic patterns, signal intensity distribution, temporal characteristics). This allows the system to distinguish between real parking spaces and false reflections by evaluating multiple signal parameters simultaneously, thereby reducing false detections while maintaining detection capability
Solution Approach 2:
The system uses feedback from the characteristic structure analysis of echo signals to continuously refine parking space detection. By analyzing the hyperbolic patterns and signal characteristics returned from objects, the system adjusts its detection criteria to differentiate between actual parking spaces and false reflections from trees or walls, improving reliability through iterative evaluation
2Device complexity
If simple distance measurement is used, then detection process is simple, but measurement precision decreases
Solution Approach 1:
The patent applies partial action by focusing analysis only on specific characteristic structures within the echo signals (hyperbolic patterns, intensity distribution in specific time windows) rather than processing the entire signal spectrum. This selective approach maintains relative simplicity while significantly improving measurement precision for parking space identification
Solution Approach 2:
The system transforms the measurement from simple distance to multi-parameter evaluation including hyperbolic structure characteristics, signal intensity distribution, and temporal patterns. This parameter transformation enables higher precision in distinguishing real parking spaces from false detections without requiring excessively complex processing
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces erroneous detections, improving the accuracy of parking space identification and allowing for adjustable sensitivity settings based on vehicle preferences or conditions, enhancing the reliability of parking space maps.
Implementation Method 1
the environmental sensor emits signals, in particular ultrasonic signals, the signals are reflected by objects in the vicinity of the vehicle and the reflected signals are received as echo signals by the distance sensor
Implementation Method 2
the reflected signals are received as echo signals by the environmental sensor
Implementation Method 3
the distance to an object, for example, can be calculated from the signal travel time
Implementation Method 4
the distance to an object, for example, can be calculated from the signal travel time
Data Source
Figure 1a~1c
Figure 2a
Figure 2b
AI summary
First, objects in the environment of the vehicle are sensed by means of at least one distance sensor of the vehicle. The reflected signals are received as echo signals by the environment sensor. In a known manner, an object distance, for example, can be calculated from the propagation time of the signals. Of the sensed objects, those that are parking-space-bounding objects are identified, for example on the basis of the spatial arrangement of said objects. Objects that bound the parking space to the side (e.g., parked vehicles) can be distinguished from lateral parking space boundaries (e.g., a curb). The identification is performed by evaluating the sensed echo signals. In the next step, the parking-space-bounding objects are classified by means of characteristic structures of the echo signals associated with the parking-space-bounding objects. Each parking-space-bounding object is assigned a certain object class. According to the invention, a parking space quality is then determined in accordance with the classification of the parking-space-bounding objects, wherein a parking space defined by the parking-space-bounding objects is then identified as a possible parking space for the vehicle if the parking space quality corresponds to a specified limit value or exceeds the specified limit value.