1. Work on an existing surveillance video camera that captures videos in the same manner as it normally does (no work needed)
2. Create a database of photos of trusted people i.e., a Trusted Persons Database or TPD (MySQL)
3. Work on a frame differencing algorithm that detects motion (OpenCV)
4. Work on a human identifier algorithm that determines if the detected moving object is a human (algorithm to be developed)
5. Determine if the identified human is moving towards the camera (algorithm to be developed)
6. Extract facial features (OpenCV)
7. Match facial features against the TPD (if a match is found then the video is not relayed to the security screen, else the video is relayed. Also, if a match is not found, then an electrified fence, until the person remains in the field of view or another person in the TPD enters the field of view and 1 millisecond has passed(Arduino/Raspberry Pi))
Note: Security cameras display live videos on a screen. As the information is not filtered and the occurrence frequency of security violation events is very low, it causes operator fatigue. Video feeds are checked only after an incident occurs and real-time preventive action is not possible. The purpose of this project is to build intelligence into the system so that it posts videos to the screen only when a potential intruder is detected.
1. are available for the work from home job/internship
2. can start the work from home job/internship between 2nd Aug'20 and 6th Sep'20
3. are available for duration of 6 months
4. have relevant skills and interests
* Women wanting to start/restart their career can also apply.
Perks
Certificate Letter of recommendation Flexible work hours
Number of openings
1
About Koyo Software Solutions
KSS is a Mumbai-based IT start-up with four lines of business, including web design, robotics, AI & machine learning, and graphics design. We make responsive, modern websites using the MERN stack and use CSS, material UI, Bootstrap, and Flexbox for styling. We are researching climbing robots. We cover all aspects of design, including functional, kinematics, dynamics, static and dynamic stability, safety, harvesting mechanism, controls, automation, and real-time communications with a remote assistant. We build solutions in Python for automated document processing and card reading for mixed, Arabic, and Latin script documents. We enable companies to better communicate their ideas or presence via a mix of graphics, videos, and animations. We design banners, visiting cards, social media backgrounds, UI/UX graphics, promotional videos, stickers, and packaging.
