PROJECT SAKHI SURAKSHA

Presented by Ditya & Mansha

Amity International School ,Noida
 ABSTRAC
T

 Safety is an inalienable right of every human being, but the safety of

women is still a challenge our world is facing today. This paper presents
smart surveillance safety tools embedded with AI capabilities like
gender detection, crowd size evaluation, and gesture recognition. This
has been developed in two stages: smart surveillance and user–side
application, including mobile apps and websites. This system triggers
SOS through gesture recognition in an emergency. It will be a user-
friendly, multilingual chatbot and community-supported system with AI
model response in minimum latency. This is a key project in which we
ensure the analysis of audio threats, geolocation, emergency
communication, real-time video streaming to capture evidence, and
such evidence are recorded automatically, sending alerts to authorities
and contacts.
 HYPOTHESIS
 We believe that if we develop a smart AI women safety system using tools like GPS,
sensors, and automatic alerts, then it will improve the safety of women and girls,
especially in emergency situations where they are not able to call or ask for help.
 Our idea is a smart device that can automatically detect danger and send a instant
message and send alerts to the control room and police station. Which will help in
We predict that:
saving time and lives.

1 2

The system will It will work in

send alerts real-time, helping
police to reach
faster than a the person
quickly.
person could do
man
 Why This Idea?

 Women’s safety is a big concern today.

Many girls feel unsafe, especially when
alone or traveling. In emergencies, they
often don’t get time to ask for help.
That’s why we thought—what if there
was a smart system that could
automatically alert someone when
danger is detected?
 This idea came from real news, cases,
and a strong wish to use technology to
protect women and make public spaces
safer.
 Purpose of the Project
 The main goal of this project is to create a smart
surveillance system that sends a quick alert to the
nearest control room , police station , NGO when
women are in trouble. It will work fast, even if the
person can't speak or use the phone.

Target Users
This project is mainly made for girls and women,
especially students and working women who may
travel alone.
 Research and Work
Done
• Studied real-life cases where women
couldn’t call for help in danger.
• Reviewed existing safety apps that need
manual action like pressing a button.
• Decided to create an automatic system
using AI, CCTV, and gesture detection.
• Added GPS and auto-alerts to send help
fast without needing phone use.
• Focused on making it easy, affordable, and
useful in public places.
 Idea
• Even with all the technology and safety steps taken, crimes against
women are still rising in many parts of the world.
• According to the World Health Organization, nearly 1 in 3 women has
faced physical or sexual violence.
• To help with this, we propose an easy-to-use smart safety system.
• It works in two ways: one uses AI-powered cameras that can detect
gender, recognize crowds, and spot SOS hand signals.
• The other is a mobile app that helps users stay safe with emergency
alerts, location sharing, and after-incident support.
• If someone makes a distress gesture, the AI model automatically
sends their location to emergency contacts and plays a loud alarm to
call for help.
 Innovation
• AI Detection: Identifies gender and distress gestures through CCTV using AI tools.
• Real-Time Action: Sends SOS, location, and alarm alerts within seconds—no delay.
• Dual Protection: Combines smart surveillance and user-friendly mobile/web apps.
• Cost-Effective: Works with existing cameras and smartphones—no extra devices
needed.
• Automatic Threat Recognition: System reacts even if the user can’t manually raise an
alarm.
• Loud Alarm Activation: Draws public attention instantly to prevent further harm.
• Direct Police Link: Sends alerts to nearby control rooms for quick intervention.
• Scalable Safety: Can be deployed in public spaces, schools, transport, and societies.
 Survey
https://docs.google.com/forms/d/e/1FAIpQLSfpreu-tL7kA1-AvRPwwtuHtnVu-KTyr15
BcEXW2OuPAGKWzQ/viewform?usp=dialog

We conducted a quick survey about women safety to gather insights and

opinions.
The responses we received have been incredibly helpful in shaping our
solution.
You can see the data analysis of responses in the upcoming slides.
It can be deployed in public spaces, schools, transport, and societies.
Data Analysis
Data Analysis
Data Analysis
Data Analysis
Data Analysis

We received around 25 responses from our survey, with

most of them filled out by women. The majority of female
respondents were under 18 and expressed that they feel
somewhat safe during the day but unsafe at night. Many
reported having seen or been aware of incidents where
women were harassed or followed in public spaces.
Public transport was commonly cited as an unsafe area,
and respondents highlighted several key factors
contributing to women's insecurity. While some
acknowledged that society occasionally takes steps
toward safety, most of them were aware of helpline
numbers and had never used them. The majority believe
stricter laws or smart surveillance could improve safety,
and while they agree technology has potential. Overall,
they believe everyone has a role to play in ensuring
women’s safety.
Diagram
Procedure
 Results

--- Expected Benchmark Results---

Face Detection: Avg Time = 0.0167s
Person Detection: Avg Time = 0.0178s
Gender Prediction: Avg Time = 0.0571s
Age Prediction: Avg Time = 0.0074s
 Cost

The basic setup for Project N

includes:
• Surveillance Camera – ₹800 to
₹1500
• Wires & Connectors – ₹200
• Power Adapter – ₹150
• Raspberry Pi (or similar processor)
– ₹3000 (optional, if needed for
smart AI features)
• Other small hardware/tools – ₹200
Estimated total cost: ₹1350 to
₹4000 (depending on the setup)
 COST EFFECTIVENESS
COST EFFECTIVENESS
• It uses affordable hardware like basic CCTV cameras
and simple wiring.
• The system can work with existing public CCTV
networks, which reduces extra setup costs.
• Once installed, it automatically detects threats, which
saves time and effort.
• It provides quick safety alerts, helping to prevent
crimes without needing constant human monitoring.
• Compared to hiring extra security or advanced
commercial systems, this is budget-friendly and
scalable.
 FUTURE SCOPE

In the future, this project can become even better

and more useful. We can add new features like we
can make a small wearable device that people can
carry easily. It can help all kinds of womens not just
women, but also children, old people, and anyone
who feels unsafe. With the help of technology and
support from others, this project can grow big and
help make the world a safer place for everyone.
 STARTUP POTENTIAL

This project can grow into a real startup. It uses AI,

cameras, and emergency alerts to improve
women’s safety in public places. It’s low-cost, easy
to install, and can work in schools, streets, or
buses. With more features like a mobile app and
real-time support, it could become a smart safety-
tech solution. With funding and guidance, it can
turn into a successful business that helps make
public spaces safer.
 LIMITATIONS

• System depends on internet – it may not

work well if the network is poor.
• CCTV cameras don’t always work in dark or
very crowded places.
• The AI might sometimes miss emergency
gestures or give a wrong alert.
• It needs electricity and constent internet
support, so can’t be used in remote or
power-cut areas easily.
• The system is still a prototype—it needs
more testing in real situations to be perfect.
 IMPROVED EFFICIENCY
• Better Cameras: Using high-resolution night
vision cameras will help in clear detection
even in low light and dark areas.
• Wireless Setup: Using wireless connections
(Wi-Fi or Bluetooth) can reduce wiring and
make the system easier to install in more
places.
• Mobile App Integration: Connecting the
system with a simple mobile app can send
alerts directly to users or security teams in
real time.
• AI Training: Training the AI model with more
real-life situations and gestures will reduce
false alarms and improve accuracy.
• Cloud Backup: Storing data in the cloud can
help in tracking patterns and keeping records
 CONCLUSION

This project was created to solve a very real and serious

problem—women's safety. Using simple technology like
cameras, wires, and artificial intelligence, we have built a smart
system that can help detect danger, send alerts, and protect
people in public spaces. It is affordable, useful, and can be
installed in many locations like schools, streets, and offices.
Through this project, we learned how technology can be used to
make the world a safer place. With some improvements and
government support, this idea can grow and help many more
people. This project is not just a school assignment—it is a small
step toward a bigger change for a safer future.
 BIBLIOGRAPHY
1.Government Reports
Highlight rise in women harassment cases in public spaces.
Emphasize need for smart surveillance in cities and transport systems.
Support use of AI and tech-based safety solutions under schemes like Nirbhaya Fund.

2. Newspaper Articles
Real stories of delayed help during emergencies.
Reports on unsafe conditions for women in schools, buses, and workplaces.
Demand for instant alert systems and panic buttons.

3. Online Research Papers

Studies proving AI can recognize gestures and facial expressions of distress.
Comparison of various safety tools (apps, wearables).
Research backing camera + AI-based surveillance for real-time action.

4. Science Textbooks (Class 8)

Concepts of sensors, circuits, and basic AI logic.
Understanding communication systems (input, processing, output).
Topics like safety, electricity, and smart technology in daily life.

5. Project Work and Observations

Camera worked well with OpenCV for motion and gesture detection.
Delay in alerts when internet was poor.
Users preferred auto-alerts over manual help buttons.
 BIBLIOGRAPHY
6. Teacher and Mentor Discussions
Suggestions on using simple tech and making the system easy for all.
Idea to add chatbot for silent emergencies.
Feedback on keeping the app child-friendly and low-cost.

7. IoT and AI Technology References

Microsoft tool used to run AI models efficiently on small devices.
IoT helps in connecting camera, app, and police station.
Low-power smart systems are possible with Raspberry Pi or Arduino.

8. Safety and Surveillance System Data

Smart cities are already using CCTVs for monitoring.Adding AI makes it proactive instead of
just recording.
Data shows fast response can save lives during attacks.

9. School Library Resources

Books on electronics, security, and women empowerment.
Case studies from past projects and competitions.
Materials on responsible tech use and social awareness.

10. Personal Interviews and News Notes

Girls shared fear of traveling alone or walking at night.
Common need: “system that works even if I can’t press a button.”
Parents and teachers showed support for smart alert systems.
THANK YOU
Presented by Mansha & Ditya (8F)
 EXAMPLES
https://justadudewhohacks.github.io/face-api.js/face_and_landmark_detection