Editor/Reviewer loginFrequency : 12 issues per year
Subject : Computer Applications and Technology
ISSN : 2319–8656 (Online)
IJCATR Volume 6 Issue 12
A Cortical Learning Movement Classification Algorithm for Video Surveillance
Abdullah Alshaikh , Mohamed Sedky
10.7753/IJCATR0612.1007
keywords : hierarchical temporal memory, cortical learning algorithms, movement classification, video forensic, post incident analysis
Classifying the movements of objects detected from a video feed is a key module to achieve a cognitive surveillance system. Machine learning techniques have been heavily proposed to solve the problem of movement classification. However, they still suffer from various limitations such as their limited ability to learn from streamed data. Recently, Hierarchical Temporal Memory (HTM) theory has introduced a new computational learning model, Cortical Learning Algorithm (CLA), inspired from the neocortex, which offers a better understanding of how our brains process temporal information. This paper proposes a novel biologically-inspired movement classification algorithm based on the HTM theory for video surveillance applications. The proposed algorithm has been tested using twenty-three videos, from VIRAT dataset, and an average accuracy of 85% was achieved.
@artical{a6122017ijcatr06121007,
Title = "A Cortical Learning Movement Classification Algorithm for Video Surveillance ",
Journal ="International Journal of Computer Applications Technology and Research(IJCATR)",
Volume = "6",
Issue ="12",
Pages ="501 - 510",
Year = "2017",
Authors ="Abdullah Alshaikh , Mohamed Sedky"}
The paper proposes a bio-inspired movement classification algorithm for video analytics
The proposed algorithm adopts the Hierarchal Temporal Memory (HTM) Theory
Cortical Learning Algorithms (CLAs) have been used to implement the proposed movement classification algorithm
VIRAT dataset has been used to evaluate the performance of the proposed algorithm.