Sunteți pe pagina 1din 16

SUBMITTED TO:

SIR ZEESHAN ASIF

SUBMITED BY:
 AROMA TARIQ 2014-AG-5744
 HARAM AZHAR 2014-AG-5740
 MAYRAJ FATIMA 2014-AG-5742
BS SE 7TH

SYSTEM FOR-
STREET CRIME
RESTRAINT
Proposal
Contents
Project Summary............................................................................................................................. 2
Organization’s History..................................................................................................................... 3
Background and Analysis ................................................................................................................ 4
Background: ................................................................................................................................ 4
Analysis: ...................................................................................................................................... 4
Proposed Goal and Implementation............................................................................................... 5
Goal ............................................................................................................................................. 5
Implementation .......................................................................................................................... 5
Functional and Non-Functional Requirements ............................................................................... 6
Functional Requirements: ........................................................................................................... 6
Non-functional requirements: .................................................................................................... 7
Project Budget ................................................................................................................................ 8
Use Case Model .............................................................................................................................. 8
Project Summary

A crowd is neither community nor society; rather, its mental attachment demonstrates
its unique group. Phenomenon in crowd could be applied in many applications. Crowd
management applications use crowd analysis to involve their management strategies to
avoid disaster caused by the crowd in public place. Public space design applications use
crowd analysis to provide guidelines for public design. Virtual environment application
models crowd as mathematic model in virtual environment to enhance the simulation of
crowd phenomena, to enrich the human life experience. Intelligent environment used
crowd analysis to assist the crowd or an individual in the crowd. Visual surveillance
application uses crowd analysis for automatic detection of anomalies and alarms. This
project relates to visual surveillance crowd analysis application.
Organization’s History
The system being made is by the students of University of Agriculture Faisalabad, Department of
computer science.
University of Agriculture Faisalabad - Overview
The University of Agriculture is a public research university in Faisalabad, Pakistan, The University
of Faisalabad is ranked 4th in Pakistan and 1st in the field of Agriculture and Veterinary Sciences
by HEC Ranking 2015. It was also ranked among top 800 world universities by QS World University
Rankings 2014.The employability ranking by Pakistan also made this institution highly reputable.
The university came among top 5 institutions of Pakistan in terms of research power.
Department of Computer Science - Overview
The Computer Science education in the University was started in 1975 in the form of short
courses with an objective to develop computer know-how. Over the years the computational
facilities were strengthened and updated to enable the University to initiate the M.Sc. Degree
programs in computer Science during the year 1992 in the department of Mathematics and
Statistics as part of its continuing computer education agenda.
In the year 1996 an independent department of Computer Science was established to perform
all the functions related to Computer Science and especially to promote and develop computer
education in the country. Currently the department is offering the degree of M.Sc. Computer
Science and M.S. Computer Science.
The curriculum of the computer Science degree programs covers all aspects of computing,
particularly Artificial intelligence, Software Engineering, Data Bases, Numerical Analysis,
Computer Architecture, Multimedia, Graphics and other rapidly evolving areas in computer
science.
The main emphasis is on project/research ranging from practical problems of application
development to deal with the abstract and philosophical issues of computer science and
advanced computing. Students are encouraged to take courses and complete a project in an area
of computer science, which interacts with Information Management, or other fields covered by
the department.
The department has excellent computing facilities available for training, data analysis, programs
development and research. These compare favorably with any university in Pakistan. The
laboratories are rich in both software and hardware installations. We hope to strengthen our
postgraduate programs, giving more emphasis on student's projects and making them more
meaningful for the software industry.
Background and Analysis
Background:
Observing humans and analyzing their actions is an active research field within the
computer vision community.
Furthermore, we only consider human detection from a single video camera but methods
that use dense stereo matching to recover the 3D information of the scene and match it
to the 3D model of humans exist.
Surveillance systems can detect and track humans from a static camera, often by
assuming the background to be static or gradually changing. These methods use a
reference image of the background only and use background subtraction to quickly detect
regions that have changed.
Another approach uses standard classification techniques to detect pedestrians in
images. A database of pedestrian/non-pedestrians images is used to train a classifier and
during run time the image is scanned at different resolutions to detect pedestrians.
Motion based methods often use the periodicity of human walk or learned gait for
pedestrian detection, which implies the use of temporal information to make a decision.
In this way, the object must be clearly visible and tracked over time. By stabilizing the
tracked object one can observe the periodic motion over time. Common to these
techniques is the assumption that each pedestrian can be properly segmented from the
background. This assumption does not always hold true.

Analysis:
Our crowd detection algorithm is based on analysis of the video sequence to control
unwanted crowd in streets. This will alarm the staff to have a glimpse of what going on
particular area which will further help them to cope with problems being arose in streets.
Proposed Goal and Implementation

Goal
The proposed system will help to restraint the increasing number of street crimes. It will be
helpful in various conditions like, it will assist the crowd density level and crowd behavior
including the number of people. Video-based crowd behavior detection aims at tackling
challenging problems such as automating and identifying changing crowd behaviors under
complex real life situations. Through extracting and integrating those crowd textures from
surveillance recordings, a redundancy wavelet transformation-based feature space can be
deployed for behavioral template matching. It will show that the abnormality appearing in crowd
scenes can be identified in a real-time fashion by the devised method. This new approach is
envisaged to facilitate a wide spectrum of crowd analysis applications through automating
current Closed-Circuit Television (CCTV)-based surveillance systems.

Our crowd detection algorithm is based on analysis of the video sequence to control unwanted
crowd in streets. This will alarm the staff to have a glimpse of what going on particular area which
will further help them to cope with problems being arose in streets.

Implementation
It can be done by using existing cameras and digital surveillance systems.

Installable in the system with cameras/server of lower performance and narrow bandwidth, as
the frame rate and video resolution required for analysis are quite lower.

It will analyze videos from surveillance cameras, quantifies crowd density level and displays
results on staff’s PC.

Display real time information of crowd and estimates wait times from the crowd density level
analyzed number of people. Staff’s PC will be updated in every (1-2) seconds.
Functional and Non-Functional Requirements

Functional Requirements:

Acquiring Crowd images


The photos should be taken with a surveillance camera. Many photos might be taken from each
camera. The software should make the difference between crowded and un-crowded pictures.
There should be some flags which the user would be instructed to look while capturing a frame.

Pre-processing image
Given crowd image, the software should alarm where the crowd is being gathered.

Detecting significance of Crowd


Given a pre-processed image, system must tell the intensity and type of crowd that whether it
is a alarming situation or not.

Finding useful measures


System must be trained in a manner that it only alerts user when it catches hustle around.

Managing users’ data records.


System must save records for later use.
Non-functional requirements:

Non-functional requirements place constraints on how the system will work, and is a
requirement that specifies criteria that can be used to judge the operation of a system , and
describe system attributes such as security , usability , reliability , efficiency , privacy and
accessibility.
In our system's Non Functional Requirements include the following:

 It should be extremely convenient to make a frame.

 For example, once the user started the system it should capture frames after every

second or as per user’s requirement the snapshot would be done and the image file

would be saved immediately. It should just need a ‘one-click’ effort.

 Instructions for acquiring crowd images should make users clear enough.

 Usability, scalability, efficiently and security.


Project Budget

Items Expenditures
Webcam\surveillance video camera 2700/= RS
Raspberry pi 4400/= RS

Total 7100/= RS

Use Case Model


Usage Scenario
Use Case ID: UC001

Use Case Login


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: User

Description: User will enter ID and password

Preconditions: 1. User must enter id and password

Post-conditions: 1. Invalid id or password

Normal Flow: 1. Access to account

Alternative Flows: Error in access

Exceptions: Internet connection problem

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil


Use Case ID: UC002

Use Case Video Input


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: System

Description: User input video

Preconditions: 2. User must enable camera

Post-conditions: 2. Camera not responding

Normal Flow: 2. Access to scene

Alternative Flows: Error in access

Exceptions: Camera connection problem with system

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil


Use Case ID: UC003

Use Case Make Frame shots


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: System

Description: System will make frames out of videos

Preconditions: 3. Videos must started

Post-conditions: 3. Frame shots saved in data base for later use

Normal Flow: 3. Frames made

Alternative Flows: Failed to make frame

Exceptions: Time constraint not mentioned

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil


Use Case ID: UC004

Use Case View unexpected crowd


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: User

Description: User will witness unexpected crowd point out by system

Preconditions: 4. System must alarm

Post-conditions: 4. User inform the relevant person

Normal Flow: 4. Crowd detected

Alternative Flows: Error in detection

Exceptions: System not trained appropriately

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil


Use Case ID: UC005

Use Case Pause surveillance camera


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: User

Description: User will pause video

Preconditions: 5. Saw unusual crowd

Post-conditions: 5. Examine the nature of crowd

Normal Flow: 5. Video paused

Alternative Flows: Video not paused

Exceptions: UI failed

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil


Use Case ID: UC006

Use Case Database Management


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: User

Description: User will manage database

Preconditions: 6. User save frames

Post-conditions: 6. Records save

Normal Flow: 6. Accessing database when needed

Alternative Flows: Not able to access database

Exceptions: Data corrupted

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil


Use Case ID: UC007

Use Case Activate Alarm mode


Name:

Date Created: 1/12/17 Date Last 10/12/17


Updated:

Actors: System

Description: System will alarm user

Preconditions: 7. System detect unusual crowd

Post-conditions: 7. Alarms user

Normal Flow: 7. Successfully notify user

Alternative Flows: Not able to notify at time of need

Exceptions: Unable to detect crowd

Business Rules: Nil

Assumptions: Nil

Notes and Issues: Nil

S-ar putea să vă placă și