This repository contains the complete code of my university project, named "Crowd detection using fractal dimension and local density". Crowd detection is the task of classifying and localizing a crowd in images.
In this project, I am particularly interested in pedestrians crowd detection. I mostly used image processing methods to isolate the crowd from non-interesting objetcs and background. Specifically, the classification task is based on fractal dimension and the localization task is based on local density.
I have written a complete article to explain the solution. The article is available at Crowd detection article
I have constructed a web application to test the proposed solution. Click here to test the Crowd Detection App
Web application :
The fractal dimension computation originates from Github repository.