Smart India Hackathon 2020

Textile Detection from CCTV Surveillance Videos

Description

The system focuses on capturing and saving various attributes of fabrics worn by targeted persons captured from various CCTV cameras through distributed intelligence along with time and location stamps over the period in a database. The database is compiled to be used in identification of suspects from video clips of crime related CCTV footages captured in a series of CCTV cameras located on routes and close to scene of crime.

Objectives

To Collect Image Dataset for Indian fabrics.
To Train our Machine Learning models for classifying the fabric labels from the video footages frame by frame.
To design a system Architecture for processing Video footages from CCTV.
To Store the Identified labels and respective timestamp in a database for the processed video footages.
To provide an user friendly interface for the operator to perform video forensics and reporting.
To Design the entire system in a more robust manner.

Pros

Responsive Web App Development
Scalable Architecture
Rich UI – UX Design
Crop Image Search
Reporting and Analytics
Dockerization of Server Script
Asynchronous and Non blocking Request Handling
High Availability and Failover Architecture
Geolocation and MapBox
Continuous Integration and Deployment Pipeline
Cross-Platform Development using Electron.js
Processing of Live feeds from CCTV
Prepared 23 Labels for Indian Fabrics
Video Quality Enhancement tools
Easy storage and retrieval of large footages and Metadata
Use of OpenSource Technologies
Speech Recognition
Use of Inhouse Developed APIs

Screenshots

Architecture

Labels

Dependencies

Installation

Prerequisites

Install Node.js, npm, Python 3.8, MongoDB and TypeScript using the link above. Follow instructions on their respecive websites. Npm is included with Node.js.
Get the API keys for Mapbox, RabbitmQ, MongoDB Cloud and fill out the environmetn variables.

Instructions

Clone the repository

git clone https://github.com/amurto/gearstalk.git

Install all the dependencies on frontend

cd frontend
npm install

Install all the dependencies on both backend servers

cd backend
npm install

cd backend_2
npm install

Usage

Run the flask server on port 5000

cd frontend
npm start

Run the frontend server on port 3000

cd frontend
npm start

Open a web browser and go to

http://localhost:3000

Contributors

Amurto Basu @amurto
Carol Sebastian @carol80
Sherwin Pillai @sherwinpillai
Mahesh Desai @mahesh131998
Shubham Bhate @ssb2920
Cassia Vaz @CassiaVaz

Mentors

Mahendra Mehra @mahendra-mehra
Elvis Dsouza @ejson03

License

MIT License Link

amurto/gearstalk