Pedestrian Crowd Detection and Segmentation using Multi-Source Feature Descriptors

Abstract

Crowd analysis is receiving much attention from research community due to its widespread importance in public safety and security. In order to automatically understand crowd dynamics, it is imperative to detect and segment crowd from the background. Crowd detection and segmentation serve as preprocessing step in most crowd analysis applications, for example, crowd tracking, behavior understanding and anomaly detection. Intuitively, the crowd regions can be extracted using background modeling or using motion cues. However, these model accumulate many false positives when the crowd is static. In this paper, we propose a novel framework that automatically detects and segments crowd by integrating appearance features from multiple sources. We evaluate our proposed framework using challenging images with varying crowd densities, camera viewpoints and pedestrian appearances. From qualitative analysis, we observe that the proposed framework work perform well by precisely segmenting crowd in complex scenes.

A Review of Deep Learning Algorithms for Anomaly Detection in Videos

“…The authors [21] present a paper on pedestrian crowd detection and segmentation using multisource feature descriptors. Input images divided into grid blocks then compute the appearance feature using multisource feature descriptor namely local binary pattern, Fourier analysis, and gray level co-occurrence matrix.…”

Internet of things based mobile application to improve citizen security

“…The results reveal that this tool contributes positively to the rationalization of the registration of police events, to the better allocation of financial and human resources and to the greater accuracy of criminal information records. Being able to identify the population and determine the exact location where a criminal act occurs is fundamental to fight crime, investigation [18] develops a novel framework that automatically detects and segments the crowd by integrating appearance features from multiple sources. It uses images with different crowd densities, camera viewpoints, and pedestrian appearances.…”