You are on page 1of 1

Faaiz Bin Nazir CHN16EC039

CROWD DENSITY INFORMATION SYSTEM IN TRAINS Rakhil P R CHN16EC080


Sandra Susan John CHN16EC089
Department Of Electronics Engineering Sundaresan T M LCHN16EC117

College of Engineering Chengannur June 2020 Guide : Prof.Dr.V.P Jyothiraj


Coordinator : Prof.Dr.Deepa J

INTRODUCTION AIM & OBJECTIVES ALGORITHM FLOWCHART


Railways provide the cheapest and most convenient mode of passenger The project aims at having a camera in each coach which constantly monitors the
transport both for long distance and suburban traffic in India for about 167 coaches and also captures images every time the train is in motion after waiting in
years.Therefore the problem of accommodating huge crowds in a single the station. After which it calculates the average number of passengers on the basis
metro train evenly in its coaches arises. Some coaches get overcrowded of the image and thereby notifies the passengers in the subsequent stops on which
compartment to get in. Thus this method is created which focuses on proximity and
while a few others remain considerably vacant while passengers board the
estimation to bring down the complexity of crowd in order to manage the
train.This project aims to solve this problem by counting the number of movement of the crowd at each layover. This thereby solves the problem of any
passengers in each coach while the metro is in transit between two stations compartment being over crowded or empty. The passengers can be notified via an
and display this count in a website, mobile app or through the information app or through the information display in the stations.
display in the stations.
METHODOLOGY
LITERATURE REVIEW
Counting people from a crowd through image is the basic function used. The CCTV
SL.NO TITLE OF THE PAPER METHODS PROPOSED ADVANTAGES AND DISADVANTAGES REMARK cameras or camera connected along the system in the train coaches can be used to
capture an image of the crowd and number of people can be estimated. This project
1 Salman Maqbool, Hasan Sajid,
“people counting in dense
A head detector is used ADV:
1.It is a crowd size estimation method
By incorporating uses YOLO algorithm. The system will detect the passengers in the coach and
to estimate spatially novel technologies,
crowd image using sparse head varying head size. After does not rely on any labelled or data of the proposed system generate a count number of the people present in the coach. The Realtime Database
detection”, 2018 crowded scenes.
sub dividing the images can be more accurate. will updates the web application periodically with the input get from the system.
into rectangular patches
SVM binary classifier 2.Requires only a generic head detector The database receives and stores the data which is number of person present in WEB USER INTERFACE
detects patch with crowd. that can be transformed by any image
Then number of head is with multiple people each bogies.
finding out by dividing
each patch by single then DISADV:
all values are summed. 1. Mainly focused for highly denser crowd. BLOCK DIAGRAM
2 Peiming Ren, Wei Fang, “A Novel People counting on ADV: Proposed system is
YOLO based real time people escalator is proposed More accurate than conventional crowd much more accurate
counting approach” , 2017 here. Camera takes the density detection. and efficient one.
picture and YOLO
SL.NO TITLE OF THE PAPER converts
METHODSimages to boxes
PROPOSED ADVANTAGES AND DISADVANTAGES REMARK
and detects objects. Real (specify open ends/future
time counting from video works)
sequence is used.

3 Prof.G.S Majuthdar ,Ashish The crowd detection is ADV: Incorporating more


Haltetal. made using infrared ray Easy to implement. efficient sensor or
“Crowd Detection in train”, 2018. sensor and Arduino kit.   camera for detection
Entering and leaving of DISADV: can improve
passengers is counted Over crowd detection is a problem performances.
and stored on data base.
Provides information on
application. REFERENCES
4 Rohit T, SaiGopal, Abhishek M. Using a camera fixed in ADV: Novel technology can
“Intelligent crowd management the compartment takes 1.It can be implemented using low cost used instead of canny CONCLUSION [1] Salman Maqbool, Hasan Sajid, “people counting in dense crowd
system in trains”, 2016 pictures at regular hardware. edge detection for
  interval converts to grey greater efficiency.
Even after 167 years of running,the railways continue to be a solid presence in the image using sparse head detection”, 2018
scale and apply canny DISADV:
edge detection. The 1.Value change for each train model is development of our nation.The cost of travel by cars and buses has seen [2] Peiming Ren, Wei Fang, “A Novel YOLO based real time people
number of edge pixel different
considerable increase,but the cost of travel by railways has more or less remained counting approach” , 2017
used for estimation of  
crowd. constant throughout.This makes the railways the backbone of our economy.Due to [3] Prof.G.S Majuthdar ,Ashish “Crowd Detection in train”, 2018.
this,there is an ever increasing use of railway services for travel and transport of [4] Rohit T, SaiGopal, Abhishek M. “Intelligent crowd management
goods.There has been numerous accidents owing to the huge rush in the system in trains”, 2016
TOOLS REQUIRED stations.These accidents occurs due to the people running to the coach which is less [5] Jia Hang Yin,Serge A Vetastian. “Image processing technique for
crowded.This project will help reduce these accidents upto some extent.By using crowd density estimation using a real time image.” , 1995
Python, OpenCV, Flask, Raspberry Pi, Camera our web app,passengers can identify the coach which is less crowded and position [6] Gauri Sabnis ,Smruti Karade. “Passenger count in metro rail
themselves accordingly,thereby reducing the risk of running and falling and getting using various video and image processing technique” , 2019
serious injuries.

You might also like