Open AccessHeadcount of the Crowd in a Congested Scene
Author(s) -
Mayur Nair
Publication year - 2021
Publication title -
türk bilgisayar ve matematik eğitimi dergisi
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.218
H-Index - 3
ISSN - 1309-4653
DOI - 10.17762/turcomat.v12i2.2331
Subject(s) - computer science , pedestrian , frame (networking) , locality , artificial intelligence , computer vision , artificial neural network , field (mathematics) , crowd psychology , density estimation , machine learning , pattern recognition (psychology) , statistics , mathematics , geography , telecommunications , linguistics , philosophy , archaeology , estimator , pure mathematics
Crowd Counting and estimation of density is really challenging and an important problem if we visually analyze the crowd. Crowd Monitoring and Analyzing Crowd behavior has been an important aspect for every research field. A lot of already existing approaches use techniques based on regression on heat maps(density) to count people present in from a single frame. These techniques however cannot restrain an individual walking and further cannot approximate the original distribution of pedestrian in the locality. Whereas, detection-based techniques detect and restrain walking men’s in the frame, but the efficiency of these techniques challenged when implemented in high-density crowd situations. To get the better of the limitations of above-mentioned problem, we have used the (Congested Scene Recognition) Neural Network. By using this type of Neural network, we are able to visualize the detection and form density map according to produce accurate outputs for the given scene. The experimental outcomes of the successfully showcases the effectiveness of the approach used.