Ultrasonic Obstacle Detection Using Echo Features and Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing obstacle detection systems using ultrasonic radars struggle with accuracy and robustness in complex driving environments due to low resolution and limited data processing capabilities, limiting their effectiveness in adapting to diverse scenarios.
Innovation Solution
Employing a machine learning model that processes ultrasonic echo data using classification and regression algorithms to enhance obstacle detection accuracy and robustness, integrating it with data preprocessing techniques such as Bayesian classification algorithms and regression algorithms, and regression algorithms to improve the accuracy and robustness of obstacle detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If ultrasonic radars are used for obstacle detection, then cost is reduced and short-range measurement capability is improved, but detection accuracy and robustness deteriorate due to low resolution data
Solution Approach 1:
The patent transforms ultrasonic echo data from time-domain to frequency-domain using Short-Time Fourier Transform (STFT), changing the parameter representation from temporal signals to spectral features. This parameter transformation enables the extraction of more informative features (frequency, amplitude, phase) that improve detection accuracy while maintaining the low cost advantage of ultrasonic radars
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between the ultrasonic radar data and the obstacle detection output. The ML model processes the transformed echo data, learning complex patterns and relationships that are not apparent through traditional methods, thereby bridging the gap between low-resolution input data and high-accuracy detection requirements
2Device complexity
If traditional ultrasonic radar methods are used, then device complexity is reduced, but adaptability to complex scenarios deteriorates
Solution Approach 1:
The patent employs a dynamic machine learning model that can adapt its parameters and decision boundaries based on the input data characteristics. The model dynamically adjusts to different scenarios (parking, driving, varying environments) by learning from training data, providing high scenario adaptability while maintaining relatively simple system architecture through software-based intelligence rather than complex hardware configurations
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
The system achieves accurate and robust obstacle detection, enabling better route planning and obstacle avoidance in complex driving scenarios while maintaining cost-effectiveness.
Implementation Method 1
obtaining ultrasonic echo data captured during vehicle movement
Implementation Method 2
obtaining ultrasonic echo data captured during vehicle movement
Data Source
Figure 1~2
Figure 3A~3C
Figure 4~5
AI summary
The present invention discloses an obstacle detection method for assisting in vehicle driving. The method comprises: obtaining ultrasonic echo data captured during vehicle movement; obtaining information associated with echo intersections based on the ultrasonic echo data; providing at least part of the ultrasonic echo data and the information associated with the echo intersections as feature data to a machine learning model to obtain detection information for an obstacle, wherein the machine learning model employs at least one of a classification algorithm or a regression algorithm; and assisting in vehicle driving based on the detection information for the obstacle.