Image credit: Unsplash Abstract Crowd density analysis is a crucial component in video surveillance mainly for security monitoring. This paper proposes a novel approach for crowd density classification, in which learned features substitute the commonly used handcrafted features.

Continue Reading