Idea Proposal Sample e Yantra Innovation Challenge 2020-21
Idea Proposal Sample e Yantra Innovation Challenge 2020-21
Idea Proposal Sample e Yantra Innovation Challenge 2020-21
Introduction/Motivation:
Potholes have been the cause of death in many accident cases as reported in various newspapers
like the Times of India, Hindustan Times etc. [1]. Every alternate day, especially during the
monsoons we get to read about how due to such bad conditions of roa ds, lives are lost. Thus
management of these potholes and road conditions has become the need of the hour.
The objective is to introduce a system that uses image processing techniques on pothole images to
identify and determine the area of potholes and provides statistical information as well as a final
geotagged image of an area with pothole location and information. The automated system would
also aid in priority based complaint attendance, having the added benefit of efficient utilization of
materials in an extremely economical way.
Hardware requirements:
1. GPS enabled camera would be required to obtain latitude and longitude information of
pothole images.
Software requirements:
1. Google Earth 7.1.5.1557 would be used for geotagging pothole images and their
Information.
2. Matlab v2007 would be used for processing images.
3. For generation of statistical data, Microsoft Excel and Google Sheets shall be used.
Implementation:
Complainant will upload pothole image on the website portal. This image will be given as input to
an automated system software that will geotag the image on the map, calculate the area, categorize
the pothole and assign a priority parameter value to it. This information will be provided to the
civil engineer of the concerned government body. The user interface for the civ il engineer will
include options for denoising, removing shadow and calculating the filling material required to
attend to the complaint.
This system is broadly divided into three parts- Geotagging, Image Processing and verification of
data acquired by a government body.
Geotagging: The map of the selected ward will be marked and its boundary will be defined using
Google Earth [6], an open source software. Using the latitudinal and longitudinal information from
the database, potholes will be geotagged onto the ward map. Description of the potholes will be
added along with its image in the geotagged part. The tagged information is stored in a kmz or kml
file that could be easily transferred via emails.
Image Processing: Problem arises when the complainant posts pothole images under varying
ambient conditions, camera resolutions, angle of photography and proximity to the pothole while
capturing the image. In images where the entire pothole is not captured, to obtain the total pothole
area, 10% of the area calculated is added to it. Affine transforms are to be applied to images where
the orientation is not in the top view. This system proposes a metric called priority parameter,
defined as: Priority parameter(ρ)= abs{P(A)*P(D) - P(AD)} where,
MATLAB software will be used in the image processing section.Steps like converting image into
black and white (binary) format,applying edge detection algorithms,using structure relation
functions to mark out the target area and filling the area with a particular distinct color to separate
it from other objects in the image will be followed in this particular sequence.
Verification of data: The third and final part of the system involves verification of data provided
by BMC to determine the precision and accuracy of the proposed system. Also priority based
pothole attendance report would be generated from this data.
Reports generated by the system will be of two types- geotag reports and statistical reports. Various
graphs and pie charts will be created using analytical tools like Google Sheets to analyze the data
that will help in determining the priority to be assigned to the repairing of the potholes and
graphically visualizing the data that was present in the theoretical form.
User Interface:
There will be separate web portals for complainants and attending officials. Complainant side web
portal would provide facilities to upload complaints and check status. Web portal for the attending
officials will provide functionalities like denoising, shadow removal option and calculate filling
material required to attend the pothole complaint.
Flow Chart:
Sequence Diagram:
Feasibility:
The current process of pothole management has certain limitations regarding coverage of all the
potholes in an area and consumption of a lot of manpower in conducting surveys to get
information about the same. In order to automate this entire process and provide a more efficient
and optimal solution, we decided to come up with the idea of a pothole management system
using image processing and geotagging. Due to the automation applied, the probability and
chances of getting optimal inferences increases.Problems like limitations in coverage of all
potholes in an area, manual calculations of material required to treat the potholes and
manipulations in costs by contractors are dealt with in an efficient manner . Transfer of
processing techniques from old paper-pen methods to completely automated methods guarantees
correct and optimal results with the added advantage of less manpower consumption.
References:
[1] Newspaper articles:
(a)http://www.mumbaimirror.com/mumbai/cover-story/Killer-pothole-leaves-Bandra-woman-in-
coma
/articleshow/46902348.cms,
(b)http://www.firstpost.com/mumbai/potholes-as-death-traps-why-should-the-civic-body-go-
scot-free
-953561.html
(c)http://indiatoday.intoday.in/story/up-engineers-contractors-to-face-culpable-homicide-
charges-if-po
tholes-cause-death/1/486135.html
(d)http://timesofindia.indiatimes.com/india/Over-11000-people-killed-by-potholes-speed-
breakers-last
-year/articleshow/48950267.cms
[2] Taehyeong Kim, Seung-Ki Ryu, “Review and Analysis of Pothole Detection Methods”,
Journal of
Emerging Trends in Computing and Information Sciences, Vol. 5, No. 8 August 2014, pp. 603-
608
[3] Christian Koch, Ioannis Brilakis, “ Pothole detection in asphalt pavement images”, Advanced
Engineering Informatics 25 (2011),pp. 507–515
[4] Emir Buza, Samir Omanovic, Alvin Huseinovic, “ Pothole Detection with Image Processing and
Spectral Clustering”, Recent Advances in Computer Science and Networking, pp. 48-53
[5] Maher I. Rajab, Mohammad H. Alawi, Mohammed A. Saif, “Application of Image Processing to
Measure Road Distresses”, WSEAS Transactions on Information Science & Applications, Issue 1,
Volume 5, January 2008, pp. 1-7
[6] Google Earth: https://drive.google.com/open?id=0B1Rifi8w3J6DMFFGeEhxX0lKQzA