Dataset, software and experimental layouts used in "Noise levels due to commercial and leisure activities in urban areas: Experimental validation of a numerical model fed with crowd density estimation using computer vision."

This repository contains the model developed in the research article entitled "Noise levels due to commercial and leisure activities in urban areas: Experimental validation of a numerical model fed with crowd density estimation using computer vision", the dataset utilised for validation and the experimental set-ups tested. The dataset includes sound level records, encompassing both sound level meter and microphone recordings, in addition to the reference source spectra employed in the model for mathematical validation. Additionally, it encompasses the outcomes of the detection of people through image analysis using computer vision, which was employed for the validation in real situation scenarios. The article proposes a strategy designed to predict the level of noise produced by crowds of people that is based on an indirect procedure that makes use of two interconnected tools: 1) an Artificial Neural Network that uses images to determine the density and distribution of the crowd in pedestrian streets and 2) a numerical model that uses this information and the urban geometry to efficiently calculate the noise level at any point in the analysis area. The model's physical foundations and the validation process, both mathematically and in a real environment, are described in the paper above, available at: https://doi.org/10.3390/s25123604