Decrease size
Reset to Default
Increase size

Intelligent Surveillance Data Retriever (ISDR) for Smart city Applications

Primary Information

Domain

Information & Communication Technology

Project No.

7794

Sanction and Project Initiation

Sanction No: F.NO.3 -18/2015 - T.S -1 (VOL.IV)

Sanction Date: 07/06/2017

Project Initiation date: 20/09/2017

Project Duration: 36

Partner Ministry/Agency/Industry

Ministry of Human Resource Development, Government of India Ministry of Urban Development, Government of India

 

Role of partner:The partner is mainly responsible for sponsoring the project. It provides the necessary review, monitoring and mentoring of the project on a periodic basis through Imprint.

 

Support from partner:Ministry of Human Resource Development, Government of India has supported 50% of the project budget. Ministry of Urban Development, Government of India has supported 50% of the project budget.

Principal Investigator

PI Image

Dr. Santos Kumar Das
National Institute of Technology Rourkela, Odisha

Host Institute

Co-PIs

PI Image

Dr. N V L Narasimha Murty
IIT Bhubaneswar, Odisha

PI Image

Mr. Chetan Kumar S
Cisco Systems India

PI Image

Mr. Amiya Kumar Samantaray
Phoenix Robotix Pvt. Ltd.

PI Image

Ms. Prangyadarsini Behera
SmartBioAccess

PI Image

Dr. Umesh Chandra Pati
National Institute of Technology Rourkela, Odisha

PI Image

Dr. Poonam Singh
National Institute of Technology Rourkela, Odisha

PI Image

Dr. Korra Sathya Babu
National Institute of Technology Rourkela, Odisha

PI Image

Dr. Santanu Sarkar
National Institute of Technology Rourkela, Odisha

PI Image

Dr. Sougata Kumar Kar
National Institute of Technology Rourkela, Odisha

 

Scope and Objectives

Building a state-of-the-art solution for smart city applications such as efficient traffic management, accident detection, object tracking, suspicious activity detection, environmental hazards, effect of pollution in real time data analysis and predictions etc. We are adding intelligence to the CCTV system by changing its embedded system architecture and making it ready for further inclusion of various sensors in-order to provide a flexible and robust platform for future smart city infrastructure. This will not only help different municipalities to take decisions driven by information, but also help in building smarter and safer cities.

Deliverables

The following modules are being delivered as part of the project. ATC (Automatic Traffic Control) AD (Accident Detection) TDMC (Traffic Density Management Control) CD (Crime Detection) PMC (Parking Monitoring & Control) CTM (Collaborative Traffic Management) EVTM (Emergency Vehicle Traffic Management) PMRC (Pollution Monitoring Reporting and Control) The items delivered as part of the project for the above modules are the following. 1. System Study (Literature Study) and Analysis Document. 2. Tools and Technology Evaluation Document 3. User Application Story Board 4. High Level Architecture Document 5. Detailed Design Document 6. Test Report 7. Product User Manual 8. Articles and Journal Publications 9. Patents

Project image

Scientific Output

An integrated system is built on a cloud platform, enabled by video analytics, sensor based technology, wireless technology, mobile computing, and big data to help provide smart traffic management & advanced surveillance system for smart city applications. ATC Automatic traffic control system is designed for providing green channel based on density of particular lane. where an IP camera is connected with video processing board, camera captured the traffic feed, from the feed vehicle classification and counting algorithm is applied and estimate the density based on number of count and area captured by camera. These density values are transmitted wirelessly to the MongoDB database and stored in respective lanes collection. In this proposed work edge computing is performed for video analysis. Traffic control panel consist of Raspberrypi as a main controller, driver unit and traffic lights, Raspberypi having capabilities of parallel processing. From the database density values retrieves and green time is calculated for particular lane based on time synchronization method and dive the signal for traffic lights through the driver unit. EVTM Emergency vehicle traffic management system provides an application for users, gives the calling facility for ambulance to reach the hospital in case of emergency. Application contains an emergency button. When you need help, our app offers an emergency button and immediately starts broadcasting your GPS location. This application wants you to turn on location services to know your GPS locations. Based on the victim's location, in case of an emergency (heart attack, accident) we are being able to broadcast the SOS to ambulance network, identify the nearest available ambulance and provide the route map driven the patient to the respective nearby registered hospital. Also provide an application for users to gives the automatic calling facility for police and fire brigade network. AD In the accident detection system we are using accelerometer, vibration, gyroscope, sound, temperature we are sending all the sensor response to the controller using wifi module. For detection of accident developed algorithm which is based on weight allocation logic. W is total weight and wth is threshold weight. Initially initialize all the sensors weight to zero. Weights will be change according to sensors outputs. If w is greater than wth there is an accident otherwise no accident. PMC In parking monitoring and control android application act as main role in software part. First, user has to register to the app by providing information like name, vehicle number, vehicle type, etc. When user login to the app, user have to choose commercial parking. In commercial parking, there are two types of parking i.e. pre booking and current booking. Pre booking shows the registered parking systems, when user books the location then app ask for vehicle type and duration of parking. App will provide the information like vehicle number, entry time, exit time, OTP, slot number, lane number, lot number, total amount. In current booking, when user reach to parking area, then in app it will show that area and also available nearby parking area. At the time of exit, by calculating the duration from entry time to exit time app will generate a bill and then notify to the user. There will be an IP Camera in entry and exit gate which will detect number plate of vehicles. We are also using ultrasonic sensors in every slots to check whether vehicle is present or not. PMRC The pollution monitoring reporting and control is a real-time pollution data acquisition platform which comprises a synergy of hardware, software and cloudware blocks. The proposed work consists of a full stack IoT architecture consisting of environmental sensors connected to a Raspberry Pi, gathering data from the physical environment and redirecting it to the cloudware using HTTP. The cloudware in this context is a NoSQL cloud database which constantly receives air quality data from the sensing nodes distributed over a vast area. Each individual sensor node measures multiple parameters viz. particulate matter (PM2.5 or PM10), carbon dioxide, carbon monoxide, oxygen, ozone, methane, ammonia, nitrous oxide, nitrogen dioxide, sulphur dioxide, hydrogen sulphide concentration, wind speed, temperature, humidity and dew-point. Based upon these concentration values, an Air Quality Index (AQI) is calculated following a dynamic time-weighted mean algorithm. At the top of the architectural stack is the user dashboard designed on Android platform which provides user-specific pollution measurement referred to as Personalised Pollution Monitoring (PPM). This PPM feature enables a user to determine the safest polluted path in an area while travelling.

 

Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image

Results and outcome till date

An integrated system is built on a cloud platform, enabled by video analytics, sensor based technology, wireless technology, mobile computing, and big data to help provide smart traffic management & advanced surveillance system for smart city applications. ATC Automatic traffic control system is designed for providing green channel based on density of particular lane. where an IP camera is connected with video processing board, camera captured the traffic feed, from the feed vehicle classification and counting algorithm is applied and estimate the density based on number of count and area captured by camera. These density values are transmitted wirelessly to the MongoDB database and stored in respective lanes collection. In this proposed work edge computing is performed for video analysis. Traffic control panel consist of Raspberrypi as a main controller, driver unit and traffic lights, Raspberypi having capabilities of parallel processing. From the database density values retrieves and green time is calculated for particular lane based on time synchronization method and dive the signal for traffic lights through the driver unit. EVTM Emergency vehicle traffic management system provides an application for users, gives the calling facility for ambulance to reach the hospital in case of emergency. Application contains an emergency button. When you need help, our app offers an emergency button and immediately starts broadcasting your GPS location. This application wants you to turn on location services to know your GPS locations. Based on the victim's location, in case of an emergency (heart attack, accident) we are being able to broadcast the SOS to ambulance network, identify the nearest available ambulance and provide the route map driven the patient to the respective nearby registered hospital. Also provide an application for users to gives the automatic calling facility for police and fire brigade network. AD In the accident detection system we are using accelerometer, vibration, gyroscope, sound, temperature we are sending all the sensor response to the controller using wifi module. For detection of accident developed algorithm which is based on weight allocation logic. W is total weight and wth is threshold weight. Initially initialize all the sensors weight to zero. Weights will be change according to sensors outputs. If w is greater than wth there is an accident otherwise no accident. PMC In parking monitoring and control android application act as main role in software part. First, user has to register to the app by providing information like name, vehicle number, vehicle type, etc. When user login to the app, user have to choose commercial parking. In commercial parking, there are two types of parking i.e. pre booking and current booking. Pre booking shows the registered parking systems, when user books the location then app ask for vehicle type and duration of parking. App will provide the information like vehicle number, entry time, exit time, OTP, slot number, lane number, lot number, total amount. In current booking, when user reach to parking area, then in app it will show that area and also available nearby parking area. At the time of exit, by calculating the duration from entry time to exit time app will generate a bill and then notify to the user. There will be an IP Camera in entry and exit gate which will detect number plate of vehicles. We are also using ultrasonic sensors in every slots to check whether vehicle is present or not. PMRC The pollution monitoring reporting and control is a real-time pollution data acquisition platform which comprises a synergy of hardware, software and cloudware blocks. The proposed work consists of a full stack IoT architecture consisting of environmental sensors connected to a Raspberry Pi, gathering data from the physical environment and redirecting it to the cloudware using HTTP. The cloudware in this context is a NoSQL cloud database which constantly receives air quality data from the sensing nodes distributed over a vast area. Each individual sensor node measures multiple parameters viz. particulate matter (PM2.5 or PM10), carbon dioxide, carbon monoxide, oxygen, ozone, methane, ammonia, nitrous oxide, nitrogen dioxide, sulphur dioxide, hydrogen sulphide concentration, wind speed, temperature, humidity and dew-point. Based upon these concentration values, an Air Quality Index (AQI) is calculated following a dynamic time-weighted mean algorithm. At the top of the architectural stack is the user dashboard designed on Android platform which provides user-specific pollution measurement referred to as Personalised Pollution Monitoring (PPM). This PPM feature enables a user to determine the safest polluted path in an area while travelling.

 

Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image
Project image

Societal benefit and impact anticipated

Whereas we have 8 key subject areas of work in Smart City, our central area of research is Automatic Traffic Control (ATC). Our system is expected to manage the traffic on its own accord based on the number of people and number of vehicles (of different type) approaching a specific traffic junction. Our electronic and cloud-based computing system is being built to automatically signal commuters through appropriate the traffic lights in a completely human-less environment and at the same time in an efficient manner than the conventional traffic system. It will help manage traffic in harsh weather conditions. It will also ensure accuracy and will be immune to human errors that may arise. It will reduce the burden of human workforce on traffic force and save the lives of officers from working in some of the unfavorable environmental conditions (scorching hot summer causing sun strokes, rainy weather, extremely cold climate etc). The system can be monitored and controlled remotely by the authorized traffic officers and relevant administrators from any location. Additionally some of the common traffic rule violation can be detected automatically and will be communicated to people and police. Illegal parking can be detected and communicated to people and police. Road safety, criminals, loitering or committing crime on the road can be detected. Traffic accidents can be detected. City pollution profiles directly or indirectly affecting traffic can be detected. Emergency vehicles e.g. Ambulance and Fire brigade vehicles can be detected automatically and necessary routing can be provided through the fastest path while taking care of the traffic. Future traffic, accidents, crime and traffic situations can be predicted. The system is expected to enhance the road traffic management, safety and security of inhabitants of smart cities.

Publications and reports

Accepted: Design of a Traffic Density Management and Control System for Smart city Applications (Conference -ICISCC-2019-106 , Bangkok) Time Series based Air Pollution Forecasting using SARIMA and Prophet Model,2019 International Conference on Information Technology and Computer Communications, ACM, Singapore (Paper ID: IT0042) Submitted: R. Nayak, M. M. Behera, U. C. Pati and S. K. Das, Video-based Real-time Intrusion Detection System using Deep-Learning for Smart City Applications, IEEE International Conference on Advanced Networks and Telecommunications Systems, Goa, India, 16-19 December, 2019

Patents

Published: ADAPTIVE HETEROGENEOUS TRAFFIC SIGNALING SYSTEM (App. No. 201931028707) In Progress: Auto Emergency Vehicle Response and Tracking System A method and system for video surveillance based automatic crime alert Multi-sensor based real time system for automatic accident detection and intimation IoT Aware Health Monitoring System Personalized Pollution management system A method for customized Safest route navigation system for smart city users

Scholars and Project Staff

1. KAILASH CH. BEHERA (3 Years) 2. R.KRISHNA CHAITANYA (3 Years) 3. RASHMIRANJAN NAYAK (3 Years) 4. GOUTAM KU SAHOO (3 Years) 5. K.KRISHNARANI SAMAL (3 Years) 6. PRASHANT DESHMUKH (3 Years) 7. LIMA PRIYADARSINI (3 Years) 8. HARSHIT SRIVASTAV (2 Years) 9. MANISHA SARANGI (3 Years) 10.PRITI TANMAYA RAY (3 Years) 11.AMULYA PARAMANIK (3 Years) 12.LAKKOJI ASHOK KUMAR (2 Years)

Challenges faced

1. Lengthy official process to buy equipment. It goes to months. 2. Sometimes the bidders don't respond to the tender and we have to keep on extending and waiting. 3. There is a defined limit on direct purchase (Rs. 25,000/ month) which creates a constraint. 4. We are not getting enough responses from Imprint to our emails requesting change in the already proposed requirements which is needed by our institute.

Other Information

We would suggest to have a dedicated point of contact in Imprint who can address the issues and concerns of the PI immediately by email and phone.

Financial Information

  • Total sanction: Rs. 38631000

  • Amount received: Rs. 24146750

  • Amount utilised for Equipment: Rs. 22686

  • Amount utilised for Manpower: Rs. 261000

  • Amount utilised for Consumables: Rs. 55761

  • Amount utilised for Contingency: Rs. 0

  • Amount utilised for Travel: Rs. 81320

  • Amount utilised for Other Expenses: 118000

  • Amount utilised for Overheads: Rs. 0

Equipment and facilities

 

PCs (14) Laptop (1) Mini UPS (18) Iron Almirah (2) Chair (25) Computer Table (16) Office Table (3) Component Self (3) CCTV Camera Set (3) NVR (2) Cisco Hub (2) Printer cum Photocpy Machine (1) Projector (1) Traffic Light Set (1) Traffic Time Counter (1) Drill Machine (1) DSLR Camera (1) Hardware Components Surveillance IP Cameras (9) Wi-Fi Router (3) Network Tool (OptiSystem Version 15.0) High-end GPU based Video Processing Board (3) Analytic Tools (Tableau Creator - 2 Users) and Tableau SW Server Licenses including AWS Cloud (5 Users) Facilities: Fully Equipped & Air conditioned ISDR Laboratory with 15 systems, modern meeting room and office room