Abstract
This article presents a system for detecting pedestrian movement patterns in urban environments, by applying computational intelligence methods for image processing and pattern detection. The proposed system is capable of processing multiple images and video sources in real-time. Furthermore, it has a flexible design, as it is based on a pipes and filters architecture that makes it easy to evaluate different computational intelligence techniques to address the subproblems involved in each stage of the process. Two main stages are implemented in the proposed system: the first stage is in charge of extracting relevant features of the processed images, by applying image processing and object tracking, and the second stage is responsible for the patterns detection. The experimental analysis of the proposed system was performed over more than 1450 problem instances, using PETS09-S2L1 videos, and the results were compared with part of the Multiple Object Tracking Challenge benchmark results. Experiments covered the two main stages of the system. Results indicate that the proposed system is competitive yet simpler than other similar software methods. Overall, this article provides the theoretical frame and a proof of concept needed for the implementation of a real-time system that takes as input a group of image sequences, extracts relevant features, and detects a set of predefined patterns. The proposed implementation is a reliable proof of the viability of building pedestrian movement pattern detection systems.
Highlights
Nowadays, there is a growing trend of installing security and surveillance cameras, with the main goal of increasing security in public spaces, streets, offices, and private facilities [1]
We proposed to develop a system based on simple image processing and computational intelligence techniques, to provide an efficient solution to be applied in real-time and realistic scenarios
This article presented a system for detecting pedestrian movement patterns, based on computational intelligence for image processing and pattern detection
Summary
There is a growing trend of installing security and surveillance cameras, with the main goal of increasing security in public spaces, streets, offices, and private facilities [1]. This article presents an approach applying computational intelligence to overcome the attentional problem of human visualizing agents that works in operational centers, making use of different techniques for image processing and pattern detection. A system capable of processing in real-time multiple image/video sources is proposed to help human visualizing agents in the process of identifying pedestrian movement patterns. Image processing and filtering techniques are applied to images generated from multiple sources, extracting relevant features of images and discarding the ones that are not of interest, according to pre-loaded rules This stage allows visualizing agents to focus their attention efficiently, on important images. We propose and study a system that conveniently combines the surveyed techniques and methods to be capable of processing multiple sources of images in real-time and detect potential events of interest.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.