Sonar Frequency Line Pattern Recognition for Vehicle Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for classifying vehicles based on their operating noise fail to automatically recognize a complete frequency line pattern within an amplitude spectrum, missing essential information such as doublets and not being operator-independent, leading to inefficiencies and potential inaccuracies in classification.
Innovation Solution
A method that automatically identifies a complete frequency line pattern by sorting and ranking frequency lines based on amplitude, forming line pairs, and determining final frequency line sets, which are then used to derive characteristic parameters for classification, allowing for operator-independent noise classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual input methods are used to enter frequency line pattern data, then the operator can control the input process, but a lot of time is lost in the classification and information may be overlooked or not entered with sufficient accuracy
Solution Approach 1:
The system automatically performs frequency line detection, sorting, and pattern recognition without requiring manual operator intervention. The computer system processes the amplitude spectrum itself to identify and classify frequency lines, eliminating the need for operators to manually enter data while maintaining high accuracy through automated algorithms
Solution Approach 2:
The manual mechanical process of operators using graphic input tools is replaced by an automated computer-based system that uses signal processing algorithms to detect and analyze frequency lines in the amplitude spectrum, substituting human manual operations with automated computational methods
2Device complexity
If only three frequency lines with the largest amplitude values are considered, then the determination process is simplified, but frequency lines with lower amplitude and individual lines within the frequency spectrum are not recorded
Solution Approach 1:
The frequency spectrum is segmented into multiple groups based on amplitude thresholds. Instead of considering only the top three frequency lines, the system divides all detectable frequency lines into different amplitude ranges and processes each segment, ensuring that both high-amplitude and low-amplitude frequency lines are captured and analyzed
Solution Approach 2:
The system performs excessive analysis by examining all frequency lines above a certain threshold rather than limiting to just three lines. This partial or excessive action ensures that no potentially important frequency information is missed, even if it increases the amount of data processed
3Ease of operation
If individual fundamental frequencies are determined based on three frequency lines, then the determination process is straightforward, but it is not possible to record a complete set of frequency lines or a complete frequency line pattern including doublets
Solution Approach 1:
An intermediary sorting step is introduced between raw frequency line detection and final pattern recognition. Frequency lines are sorted by amplitude and processed through multiple analysis stages, allowing the system to build a complete frequency line pattern that includes fundamental frequencies, harmonics, and doublets while maintaining operational simplicity through structured processing
Solution Approach 2:
Frequency lines are preliminarily sorted and ranked by amplitude before the main pattern recognition process. This preliminary organization of data prepares the information in advance, making the subsequent pattern detection more accurate and complete without complicating the overall operation
Data Source
Figure 1
Figure 2~3
AI summary
The invention relates to a method for determining a frequency line pattern within at least one amplitude spectrum that is produced from audible signals that have been radiated by at least one detected vehicle and received by means of a sonar installation. First of all, a predetermined number of frequency lines in the or each amplitude spectrum are selected and are sorted in order of decreasing amplitude magnitude and are selected using a threshold value that ensures that a predetermined suitable number is not exceeded. Following stipulation of a plurality of threshold values, frequency lines that have been selected taking account of at least one threshold value are compared with one another. The line pairs formed by the comparison are used to determine preliminary frequency line sets with an associated fundamental frequency that are rated on the basis of their identifiable structures. An exclusion method is used to form the final frequency line sets, taking account of a rating that is dependent on the sequence, the number, the density, the amplitude and the fundamental frequency of the frequency lines. It is then possible to find strong single lines that the spectrum may contain. For each detected vehicle, it is therefore possible to assign a target-oriented frequency line pattern that can be used to classify the vehicle. The invention also relates to an appropriate apparatus for carrying out the method according to the invention.