HUMAN MOTION DETECTION AND TRACKING FOR VIDEO SURVEILLANCE

HUMAN MOTION DETECTION AND TRACKING FOR VIDEO SURVEILLANCE

DOMAIN INTRODUCTION:

Image processing is any form of signal processing for which the

input is an image, such as a photograph or video frame; the output of image

processing may be either an image or a set of characteristics or parameters

related to the image.

Video content analysis (also Video content analytics, VCA) is the

capability of automatically analyzing video to detect and determine temporal

events not based on a single image. As such, it can be seen as the automated

equivalent of the biological visual cortex.

ABSTRACT:

An approach to detect and track groups of people in video-surveillance

applications, and to automatically recognize their behavior.

This method keeps track of individuals moving together by maintaining a

spacial and temporal group coherence.

First, people are individually detected and tracked. Second, their trajectories

are analyzed over a temporal window and clustered using the Mean-Shift

algorithm.

A coherence value describes how well a set of people can be described as a

group. Furthermore, we propose a formal event description language.

The group events recognition approach is successfully validated on 4 camera

views from 3 datasets: an airport, a subway, a shopping center corridor and an

entrance hall.

EXISTING SYSTEM: Existing event modeling techniques in three categories: the pattern recognition models, the state based models and the semantic models. First category is artificial vision, which tries to extract visual characteristics and identify objects and patterns. A second option is to reuse existing metadata and try to enhance it in a semantic way. Finally, using the combined result of collaborative human efforts can lead to data that is otherwise difficult or impossible to obtain.

PROPOSED SYSTEM:

Detect and track groups of people in video-surveillance applications, and to automatically recognize their behavior. In the framework of a video understanding system video sequences are abstracted in physical objects :objects of interest for a given application. Then the physical objects are used to recognize events. The proposed event detection approach correctly recognizes events but shows its limitation for some specific events (e.g. fighting is best characterized by internal group movement).

SYSTEM REQUIREMENTS:

Hardware System Configuration:-

Processor - Pentium –IV

Speed - 1.1 Ghz

RAM - 512 MB(min)

Hard Disk - 40 GB

Key Board - Standard Windows Keyboard

Mouse - Two or Three Button Mouse

Monitor - 15” Samsung Color Monitor

Software System Configuration:-Operating System - Windows XP/7/8

Framework - Visual Studio 2008

FrontEnd - C#.NET

MODULES DESCRIPTION:

Video File:

Video File is going to be the input for the system. First we need to upload the

input video file which contains human movement activity such as chatting,

walking together, etc...

Then we can to detect the human activity appropriately.

Foreground Blobs Detection:

Blobs of foreground pixels are grouped to form physical objects classified

into predefined categories based on the 3D size of objects group of persons,

person and noise.

When people overlap or are too close to each other, segmentation fails to

split them and they are detected as a single object classified as group of persons

because its size is bigger than the size of a single person.

Those classes of objects are specified using gaussian functions.

MODULES DESCRIPTION:

Physical Object Tracking:

Video sequences are abstracted in physical objects: objects of interest for a

given application.

Then the physical objects are used to recognize events before tracking group

events.

Group Tracking:

Group tracking is based on people detection. The people detection can be

performed by various methods.

For group behavior recognition, detected group objects within the video

sequence and scene context objects (zone, equipment) are described.

The scene context objects help to recognize specific events.

MODULES DESCRIPTION:

Event Detection:

Event recognition is a key task in automatic understanding of video sequences.

The typical detection algorithm takes as input a video sequence and extracts

interesting objects (physical objects). Then, these objects of interest are used to

model events.

Finally, the events are recognized. The abstraction stage determines which

modeling techniques can be applied.

The output of the group tracker, which is the input of the event detection, is a set

of tracked groups (keeping a consistent id through frames) having properties (such as

the intra-objects distance) and composed of detected physical objects at each frame.

SCREENSHOTSMENUS:

FILE MENU:

SCREENSHOTSOPEN TAB:

PLAYING VIDEO FILE:

SCREENSHOTSEVENT DETECTED:

SCREENSHOTSEVENT DETECTED:

SCREENSHOTSADMIN LOGIN:

ADMIN MENU:

SCREENSHOTSMANAGE USER TAB:

POLICE MOBILE NUMBERS TAB:

SCREENSHOTSMOTION MENU:

GRAPH MENU:

CONCLUSION:

We propose a generic, plug and play framework for event recognition from videos

The scientific community can share a common ontology composed of event models

and vision primitives.

We demonstrate this framework on group behavior recognition applications, using a

novel group tracking approach.

This approach gives satisfying results even on very challenging datasets (numerous

occlusions and long duration sequences) such as in figure 6.

The vision primitives are based on global attributes of groups (position, speed, size).

The proposed human event detection approach correctly recognizes events but shows

its limitation for some specific events (e.g. fighting is best characterized by internal

group movement).

Moreover, in this work the gap between video data and semantical events is modeled

manually by vision experts, the next step is to learn automatically the vision primitives.

FUTURE ENHANCEMENT:

The primary aim of this research is to develop a framework for an automatic semantic

content extraction system for videos which can be utilized in various areas, such as

surveillance, sport events, and news video applications.

First of all, the semantic content extraction process is done automatically. In addition,

a generic ontology-based semantic met ontology model for videos (VISCOM) is

proposed.

An automatic Genetic Algorithm-based object extraction method is integrated to the

proposed system to capture semantic content.

In every component of the framework, ontology-based modeling and extraction

capabilities are used.

The test results clearly show the success of the developed system.

As a further study, one can improve the model and the extraction capabilities of the

framework for spatial relation extraction by considering the viewing angle of camera and

the motions in the depth dimension.

REFERENCE:TEXT BOOKS:F. Bobick, J.W. Davis, I. C. Society, and I. C. Society. The recognition of human movement using temporal templates.D. P. Chau, F. Bremond, and M. Thonnat. A multi-feature tracking algorithm enabling adaptation to context variations.X. Chen and C. Zhang. An interactive semantic video mining and retrieval platform–application in transportation surveillance video for incident detection.E. Corv´ee and F. Bremond. Haar like and LBP based features for face, head and people detection in video sequences. In IWBAVU (ICVS 2011), page 10, Sept. 2011.T. V. Duong, H. H. Bui, D. Q. Phung, and S. Venkatesh. Activity recognition and abnormality detection with the switching hidden semi-markov model. WEB REFERENCES: http://www.microsoftvirtualacademy.com/training-courses/c-fundamentals-for-absolute

-beginners. http://C#snippets.com/Articles/Simple-User-Registration-Form-Example-in-CSharpNet

.aspx. http://www.vijaymukhi.com/documents/books/csbasics/csharp1.html. http://www.networkcomputing.com/.