Group members - Reena Kale - 23 TE COMPS

Yash Shah - 53
Brian Pereira - 71

REPORT TOPIC: Things in the air: tagging wearable IoT

information on drone videos

PROBLEM STATEMENT -
‘Things in the air: tagging wearable IoT information on drone videos’.
Drones have been applied to a wide range of security and surveillance
applications recently. With drones, Internet of Things are extending to 3D
space.Pid cant be used with drones because it requires a close contact such
as RFID or fingerprint/iris/face recognition.This tagging can be used in
Unmanned Aerial Vehicles, UAVs,military,construction site monitoring and
environmental monitoring. It helps in location tracking person tagging which is
very helpful in military and construction site monitoring where people are
required to wear badges so whenever needed, the fusion server can also
send alert messages to specific persons (for reasons such as the persons
entering a dangerous region found by the drone).They want PID in an aerial
platform.

TECHNOLOGY-
1. Drone
2. Deep learning
3. Data fusion
4. YOLO (for human tracking)
5. DISCRETE WAVELET TRANSFORM algorithm
6. Aruco markers
7. Tracking through SORT

PROBLEM SOLVED -
First of all, the drone records the video.The video of drone uses You Only
Look Once (YOLO) to retrieve all objects that are recognized as human.Using
tagging IoT information on human objects in a drone view as an example.
This allows them to visualize not only people’s IDs, but also their personal
profiles.If a person is not willing to share then he is marked as “unknown
persons”.

The coordinate system must match the real world so they placed the aruco
markers that transform a pixel location to a global space background with
respect to markers location.
An ArUco marker is a synthetic square composed by a wide black border and
an inner binary matrix which determines its identification.Assuming assume
that our wearable IoT devices have built-in inertial sensors, including
accelerometer, gyro and magnetometer field sensors.There are a number of
people walking in the square ground, all putting on wearable devices on their
chests (phones).
They retrieve all human objects from each image by YOLO and then establish
frame-to-frame association to connect these human objects by SORT. It is
applied to associate the same person in adjacent frames. However, when a
person leaves the camera view or is partially occluded, his ID is destroyed. If
he re-enters the view, a new unique ID will be created.

After having the video data fusion is applied.

Video feature extraction: Motion features such as rotation, direction, behavior,
and movement of the trajectory.
Sensor feature extraction: For fusion purpose,they needed to retrieve the
same features from wearable sensor data as those from videos.
Data fusion calculates the correlation of video and data sensors.The main
issue they faced was the time delay faced in the synchronization between the
wearables and drone.As these things might work on their own clocks and
sampling frequencies they used Discrete Wavelet Transform algorithm which
generates a similarity matrix indexed for each pair.
The other problem was of data loss as the communication was wireless.

Any questions:
a) What are the advantages and disadvantages of IOT?
b) What is the connection between IoT and sensors in the commercial
enterprise?
c) Why is there a direct and reverse action on a controller when the control
valves are already having direct /reverse actions ?
d) The people who shared the personal information is safe(not shared or
leaked or hacked)?

Conclusion drawn:To conclude, our results show an interesting new

direction that future human identification is not limited to face recognition; it is
also possible to use IoT for human identification. Furthermore, in addition to
human identity, we are able to provide personal profiles/preferences/hobbies
from IoT devices as well. This would make future surveillance applications
even more informative.The proposed PID and tracking system requires the
authorization of accessing. IoT data has  many applications in military training
and construction site monitoring where people are required to wear badges
and they have mixed drone camera and initial data from wearable devices for
person identification and tracking . They achieved person identification and
tracking by changing drone angles which did not rely on human biological
features.