Density Based Crowd Couting is getting used more often to further improve the work of various areas in modern society. State-of-the-art methods accomplish better results estimating counts and positions of people, every day which leads to the increase usage of such methods. The main goal of this thesis is to give a vast overview of current state-of-the-art methods for Density Based Crowd Counting. Apart from that, an own approach to solve this problem is introduced and evaluated. This approach tries to further minimize the estimation error in Crowd Counts and corresponding Density Maps. The evaluations results depicts, that the own attempt still lacks behind modern methods in quite a few aspects. Despite this, it also improves the ability of state-of-the-art methods to estimate better Density Maps of Data the network has not seen before. Further resurch has to be conducted to analyse the underlying idea further improving accomplished results.
Master of Science in Computer Science. Full-time nerd. Faible for bad puns.