Automatic Detection and Tracking of Pedestrians in Videos with Various Crowd Densities

Dehghan, Afshin; Idrees, Haroon; Zamir, Amir; Shah, Mubarak

doi:10.1007/978-3-319-02447-9_1

Cited by 29 publications

(11 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To reduce all experiments to a common denominator, we implemented a base movement tracking algorithm (BMTA) based on OpenCV corner detection and optical flow analysis. This algorithm manipulates all pixels on each video frame using the goodFeaturesToTrack [ 36 , 37 , 43 , 44 ] function plus a few algorithmic optimization presented in [ 11 , 30 ]. We use this tool as a reference point for all comparisons.…”

Section: Resultsmentioning

confidence: 99%

“…The number of surveillance cameras in urban areas has been increasing at a rate which results in massive amounts of video to be analyzed. As a consequence, in the last decade, many approaches [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 ] have been proposed for moving object detection and tracking from videos. Researchers concentrate on traffic monitoring and security, visual surveillance, and sensors networks able to analyze complex area and flow movements.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Moving Object Detection in Heterogeneous Conditions in Embedded Systems

Garbo

Quer

2017

Sensors

View full text Add to dashboard Cite

This paper presents a system for moving object exposure, focusing on pedestrian detection, in external, unfriendly, and heterogeneous environments. The system manipulates and accurately merges information coming from subsequent video frames, making small computational efforts in each single frame. Its main characterizing feature is to combine several well-known movement detection and tracking techniques, and to orchestrate them in a smart way to obtain good results in diversified scenarios. It uses dynamically adjusted thresholds to characterize different regions of interest, and it also adopts techniques to efficiently track movements, and detect and correct false positives. Accuracy and reliability mainly depend on the overall receipt, i.e., on how the software system is designed and implemented, on how the different algorithmic phases communicate information and collaborate with each other, and on how concurrency is organized. The application is specifically designed to work with inexpensive hardware devices, such as off-the-shelf video cameras and small embedded computational units, eventually forming an intelligent urban grid. As a matter of fact, the major contribution of the paper is the presentation of a tool for real-time applications in embedded devices with finite computational (time and memory) resources. We run experimental results on several video sequences (both home-made and publicly available), showing the robustness and accuracy of the overall detection strategy. Comparisons with state-of-the-art strategies show that our application has similar tracking accuracy but much higher frame-per-second rates.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Moving Object Detection in Heterogeneous Conditions in Embedded Systems

Garbo

Quer

2017

Sensors

View full text Add to dashboard Cite

show abstract

“…In a public place, for instance, it is necessary to understand how 'objects move according to constraints imposed by: (1) the environment (highway, inner city); (2) the type of interactions they are exposed to (cruise, turning); and (3) their nature or class category (car, pedestrian)' (Romero-Cano, Agamennoni & Nieto 2016: 654). Pedestrian detection can improve understanding of the movement of people in special situations (Dehghan et al 2013) as well as agent-based modelling to visualise and understand pedestrians' behaviour in order to improve planning processes such as those for the new London Bridge national railway station (Le Glatin, Milford & Hutton 2014).…”

Section: Permanently Installed Sensorsmentioning

confidence: 99%

Defining and assessing walkability: a concept for an integrated approach using surveys, biosensors and geospatial analysis

Dörrzapf

Kovács-Györi

Resch

et al. 2019

Urban Development Issues

View full text Add to dashboard Cite

Walking as a transport mode is still often underrepresented in the overall transport system. Consequently, pedestrian mobility is usually not recorded statistically in the same manner as it is performed for motorised traffic which leads to an underestimation of its importance and positive effects on people and cities. However, the integration of walkability assessments is potentially a valuable complement in urban planning processes through considering important quantitative and qualitative aspects of walking in cities. Recent literature shows a variety of approaches involving discrepancies in the definition of walkability, the factors which contribute to it, and methods of assessing them. This paper provides a new understanding of the concept of walkability in the European context. Our approach relies on the extension of methodological competence in transportation, spatial planning and geography by linking new measurement methods for evaluating walkability. We propose an integrated approach to assessing walkability in a comprehensive methodology that combines existing qualitative and GIS-based methods with biosensor technologies and thus captures the perceptions and emotions of pedestrians. This results in an increased plausibility and relevance of the results of walkability analysis by considering the spatial environment and its effect on people.

show abstract

“…They experimented with these classifiers and algorithms and produce an output which tracks human heads through occlusions. Afshin Dehghan et al [8] stated two methods for pedestrian tracking in videos with different crowd density. For videos with low density, each person is first detected using a part-based human detector.…”

Section: Previous Workmentioning

confidence: 99%

A Novel Approach to Detect Crimes and Assist Law Enforcement Agency using Deep Learning with CCTVs and Drones

Pawar¹

2019

IJRASET

View full text Add to dashboard Cite

In India, crime has always been a serious issue and it is strenuous to tackle it due to lack of technology. The current security techniques are precarious and cannot be trusted, due to the negligence of the authorities and the policy setters, there's a monumental rise in the number of crimes taking place daily like kidnapping, abduction, molestation, etc. India ranks 133 out of 167 countries in Women, Peace and Security Index with Indore reporting the highest crime rate (769.1) among the megacities in India, followed by Bhopal (719.5) and Jaipur (597.1). Thus, an improved security system near bus-stops, schools, neighborhoods, and other public places is mandatory to detect violent or abnormal activities to avoid any casualties which could cause social or economic damage. The main focus should be to aid law enforcement agencies to avert the outbreak of violence by monitoring crime scenes and alerting the respective security officials. There is a need to instigate fear and perturbation in the minds of the criminals that they will get caught on committing any type of crime. Several image-based extraction techniques with smart methods of deep learning and artificial intelligence will help support the security officials to detect, monitor and track the perpetrator. Thus, we present a smart way to reduce violence, by criminal activity detection and recognition in a particular CCTV's field of view and make the drone identify, track and chase the criminal till he or she gets apprehended.

show abstract

Automatic Detection and Tracking of Pedestrians in Videos with Various Crowd Densities

Cited by 29 publications

References 16 publications

Moving Object Detection in Heterogeneous Conditions in Embedded Systems

Moving Object Detection in Heterogeneous Conditions in Embedded Systems

Defining and assessing walkability: a concept for an integrated approach using surveys, biosensors and geospatial analysis

A Novel Approach to Detect Crimes and Assist Law Enforcement Agency using Deep Learning with CCTVs and Drones

Contact Info

Product

Resources

About