Convolutional Neural Network and LSTM Applied To Abnormal Behaviour Detection From Highway Footage

• DOI: 10.14210/cotb.v13.p051-058
• Publication Date: 2022-07-15T09:42:17Z
• Authors: Rafael Marinho de...; Elcio Hideti Shig...; Rafael Duarte Coe...
• Keywords: SAFER; Tracking (education)

ABSTRACT Relying on computer vision, many clever things are possible in order to make the world safer and optimized resource management, especially considering time attention as manageable resources, once modern is very abundant cameras from inside our pockets above heads while crossing streets. Thus, automated solutions based vision techniques detect, react or even prevent relevant events such robbery, car crashes traffic jams can be accomplished implemented for sake of both logistical surveillance improvements. In this paper, we present an approach vehicles’ abnormal behaviours detection highway footages, which vectorial data displacement extracted directly footage through object tracking with a deep convolutional neural network inserted into long-short term memory behaviour classification. The results show that classifications consistent same principles may applied other trackable objects scenarios well.