Calibration method for automobile whistling snapshot device based on deep learning

Abstract

The invention relates to a calibration method and system for an automobile whistling snapshot device based on deep learning. The calibration system comprises an automobile whistling snapshot device,a sound source, a black and white grid disc, a transmission device, a signal generation device and a computer. The automobile whistling snapshot device is fixedly installed at the position right opposite to the sound source and is connected with a computer for processing data. The sound source is loaded on a transmission device consisting of a plurality of sections of guide rails, and a black andwhite grid disc is fixed right behind the sound source; the sound source is connected with the signal generating device and used for simulating different automobile whistling sounds. By capturing a sound source, a pseudo-color image and a black and white grid disc in the acoustic phase cloud image generated by the whistling snapshot device through the neural network, the positioning precision andthe resolution precision are acquired by calculations. The performance of the whistling snapshot device is rapidly detected in real time, the product research and development quality are ensured, andthe performance index requirements of enterprises for research and development products are improved.

Description

technical field [0001] The invention relates to the fields of machine learning and acoustic measurement, in particular to a calibration method for a car whistle capture device based on deep learning. Background technique [0002] As the number of motor vehicles owned by urban residents continues to increase year by year, the pressure on urban traffic is also increasing, and traffic pollution has become a problem that cannot be ignored. As an important part of traffic pollution, car horn noise has had a huge impact on people's lives. Relevant law enforcement departments in my country have also promulgated relevant laws to limit or prohibit the sound of car horns on certain road sections. Traffic police departments often use patrols to correct violations and other methods to supervise car horns. However, manual law enforcement is not only inefficient, but also wastes a lot of time. Therefore, how to efficiently locate and identify the illegal behavior of honking in urban traff...

AI Technical Summary

Problems solved by technology

[0004] At present, the only sound source localization calibration specification in China - "Sound Source Identification and Positioning System (Beamforming Method) Calibration Specification" initially provides technical standards for the quantitative evaluation of sound source localization systems, but the calibration specification is only applicable to beamforming algorithms sound source localization system, and the calibration process described is limited to indoor calibration, and no corresponding regulations are made for on-site calibration

Images