Smart surveillance security system detecting weapons.

Project Title

Smart surveillance security system detecting weapons.

Project Area of Specialization Artificial IntelligenceProject Summary

Man is prone to error but when there comes the human life, we can’t neglect the iota of negligence and carelessness. According to statistics 4.20 in 100000 inhabitants are killed in mass shooting in Pakistan every year. We aim to design an embedded system for surveillance security that detects the weapons in real time, if detection is positively true the system prompts the security personals to tackle the situation by reaching at the point of incident through IP cameras rapidly. A model that can 1). alert the human operator when a firearm or gun is visible in the frame, 2). Automatic doors locking system when shooter appears caring heinous weapon. So, for the real-time detection we intent to achieve the high computational power from raspberry pi with 2 to 4 GB RAM and at least 720p camera module where 60 frames per seconds are enough to feed the model, subsequently if the gun is detected in the camera our proposed model will check if it is active shooter then with in no time it will prompt the Adriano to generate the signal for alarm. If possible, through IP webcams we can also share the live photoaged to near security personals to take the action in mean time. Additionally, we will attach the database for recording all the activities in order to carry effect actions in the metropolitan areas for future emergency.

Project Objectives

We aim to make a smart high computational circuitry which will help us to first build custom object detection here our objects of interest are guns later knives can be included too, we will use Raspberry camera module 720p for training and testing data after training the YOLO/CNN/HAAR cascade classifier we will implement programming on raspberry pi for real-time detection in the end we will have an embedded system that will notify security personals through Arduino for alarming to take rapid response through IP cameras.

• We aim to make our own dataset of guns and knives using Raspberry camera module.
• For real-time weapon detection we will implement the programming on Raspberry pi.
• If the detection of active shooter is positively true, the system will generate the signals using Arduino to security personals to take actions through IP cameras.
• The latitude, longitude of the place of incident will be fetched to near by police/security personals for rapid response.
• Date, timing and pictures of active shooters and place of incident will be saved in database in case Government could take effective actions in future.
• We aim to design an auto-lock system too, if the weapon is detected in surveillance here people can be saved from active shooter and shooter can be caught with red handed.

Project Implementation Method

We aim to make own dataset containing images of weapons through Raspberry pi camera module 720p for training and testing of model. After exporting the trained model, we will program the Raspberry pi with our model to detect the weapons in real time. If the active shooter caring gun appears in the live frame the circuit will generate signal to Arduino, from there through Wi-Fi module we will send multiple signals to buzzers, LEDs and screen if possible, a voice over telling the live location of the incident. Here we have two possibilities,

1). Positive true:  When gun appears in surveillance and it notifies the system to take action.

2). Negative true: When gun doesn’t appear in surveillance and it alerts the system, it would be reduced as much as possible.

Benefits of the Project

We have lost thousands of crucial lives in mass shooting in all our the world from Crist Church Mosque (killing 50+) incident to Army Public School Peshawar (killing 150+) and this list goes on and on. It is common fact that if a person watches same thing for hours, there is chance of laziness and carelessness here it is same case with security guard/watchman who is keeping eye on every scene behind the CCTVs. On the one hand our proposed system ease the job of watchman and help common people to escape from the dangerous situation about to happen and secondly if escape is not possible as in plazas, the doors will be auto-locked as a culprit may not harm the people and in mean time security personals will be informed too about the scenario in real-time. Precisely, if we adopt this model in our surveillance system, we can achieve the drastic change in homicide/mass murdering. Applying this technology, we can decrease the overall demise ratio of our common people and make the society more peaceful. The core focus of the project is smart security from the hazardous active weapons, where there is surveillance there should be an active eye keep tracking and alerting the concerned people to take speedy actions.

Technical Details of Final Deliverable

The final product is an embedded system where the Raspberry pi is programmed to detect the active shooters caring weapons if we get 100% positively true results in the end, we’ll extend the scope of dataset to knives and other damaging weapons. The Arduino and wifi is responsible to notify the buzzers, LEDs and other screens to show the place of incident, meanwhile the designed circuitry will prompt the nearer security squad to take rapid response then and there through IP cameras as well. In case if the distance is too long or some other hindrances in the path the doors at the place of incident will be auto locked to protect the people from homicide/ robbing. Additionally, we are attaching database as well to record all the activity in order to take effect actions in future.

Final Deliverable of the Project HW/SW integrated systemCore Industry SecurityOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Industry, Innovation and Infrastructure, Sustainable Cities and CommunitiesRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	23075
Raspberry PI 4 Module	Equipment	1	15000	15000
Raspberrry PI camera module	Equipment	1	5500	5500
Arduino Mega 2560	Equipment	1	1475	1475
Esp8266 Cp2102 Wifi Module	Equipment	1	600	600
Breadboard	Miscellaneous	2	150	300
Connecting Wires	Miscellaneous	40	5	200