SECURITY ALARM FOR BANK AND MUSEUM
The use of security cameras is no new to us. But these have had been carrying a significant flaw that they were not capable enough to differentiate a human from an object. Only detecting a moving object is not enough as the device should be able to tell what kind of thing it is and should be more pr
2025-06-28 16:34:55 - Adil Khan
SECURITY ALARM FOR BANK AND MUSEUM
Project Area of Specialization Artificial IntelligenceProject SummaryThe use of security cameras is no new to us. But these have had been carrying a significant flaw that they were not capable enough to differentiate a human from an object. Only detecting a moving object is not enough as the device should be able to tell what kind of thing it is and should be more precise. Finally, AI has been successful in overcoming that drawback through machine learning. Now, it can not only know it’s a human, but it can also detect abnormal or suspicious behaviour and ring the alarm for you. The goal is to merge AI and machine learning together and train it to expert level to build a safe and secure atmosphere for you. To add to this, we can take care of the data dependency by linking it with another device. Hence, our contribution to the security alarm and camera is that by integrating AI, we have minimized the chances of false alarm generation, which were relatively high previously as we are not going on random motion detection and has made the model quite specific to ensure a safer environment for all of us. Taking museums as an example, they are entirely responsible for securing and also exhibiting the priceless treasures which require expert security systems. To further enhance the model, the addition of necessary measures like infrared signals, audible signals, motion sensors etc., is essential.
Project ObjectivesMinimize Crime Rate
Provide Constant Protection
Increases Security at Banks and Museums
Provide Uninterrupted Functional System
Provide Secure System
Provide Theft Detection with Special Alarm
Project Implementation MethodIt is layered and intertwined between the infrastructure and the algorithm. The hardware part has four dimensions; a processor, GPU, special chip, and a highspeed network. All of them work in various ways, and the particular chip is solely dedicated to the AI algorithm. The framework used for the algorithmic layer of this specification is TensorFlow. It permits the apt algorithms of AI and runs them. The libraries in TensorFlow cover all the necessary methods that we need in this aspect. It is developed by Google and has been very useful in the AI fieldwork. Next comes the machine learning aspect of AI technology. We opted for AI in the first place because we wanted the machines or computer systems to make human-like decisions and results on its own by providing it with accurate and sufficient data. Hence, our model too, has acquired the needed training to calculate new data. For example, the human detection system through machine learning is responsible for detecting objects and coming to decide on its own about whether or not an item is a person. Hence, the human-like tasks are now handed over to such systems and have been working quite well. It has also reduced human error chances. The accuracy of such systems gets based upon the machine learning part. Enough data is likely to promote more accurate results. The functionality of AI on its platforms offers complete AI functions, including graphical outlooks, data analysis, sharing capabilities, semantics and other such functions. Also, keep in mind that the learning process is supervised and is of great importance in this niche.
Benefits of the ProjectPrevent Robberies
Integration With Alarm Systems
PROTECTS VALUABLES
HELPS SAVE ENERGY
HOME AUTOMATION CONVENIENCE
LOWERS INSURANCE
CONTROLLED SURVEILLANCE
Technical Details of Final DeliverableIn this the final deliverables are as a product form that it would have raspberry pi 4 with multiple ip cameras which would run simultaneously and we will run each camera alarm separately. All ip cameras would be connected with raspberry pi 4 and set alarm for human detection and the person detection and it would run separately for each webcam in bank and museums. It would become one go security system for any bank, museums, office, building etc. If we talk about home security so it would be also applicable in this particular scenario. Core technical intelligence technology can be divided into a bottom-up algorithm layer and an infrastructure layer. The structure includes some basic functions, including a processor, GPU, a high-speed network and a dedicated AI chip. We can build a framework for algorithms such as Tensorflow, Caffe, imutils, numpad and other built-in libraries based on this basic hardware. Above the infrastructure layer is the algorithm layer. The most common algorithm layers are engineering algorithms, including a series of engineering algorithms, such as in-depth training, transport training, general opposing networks, and reinforcement training. The AI-based model has its design on python along with other necessary libraries. The model is a trained one and gets presorted data and hence, can work for object detection. Though the raspberry pi 4 is an excellent handler for our model, it might get heated up at a certain point. But it can restore its power and continues working when the temperature is back to normal. To break down the functionality of the model for you, when the model gets live streaming footage of an event, it breaks the video down into several layers which then refines itself following the resolution power that we have. Since the systems only know numbers, every factor of the imaging process matters. The pixels then get simplified into parameters which go down to patterns. Every step defined here is a part of the machine learning phase. After the breaking down of these components, we reach one resultant node, which is the decision-making statement that decides whether the object is a person or not. That is how the detection system works.
Final Deliverable of the Project HW/SW integrated systemCore Industry SecurityOther Industries Education , IT , Medical , Manufacturing , Telecommunication Core Technology Artificial Intelligence(AI)Other Technologies Internet of Things (IoT)Sustainable Development Goals Industry, Innovation and InfrastructureRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 44550 | |||
| Raspberry Pi | Equipment | 2 | 14500 | 29000 |
| Pi Box | Equipment | 1 | 1000 | 1000 |
| Pi Fan | Equipment | 1 | 250 | 250 |
| Adpter | Equipment | 2 | 650 | 1300 |
| 32Gb Memory Card | Equipment | 2 | 1100 | 2200 |
| HDMI Cable | Equipment | 1 | 300 | 300 |
| Alarm Burzer | Equipment | 1 | 500 | 500 |
| Stationary-Books | Miscellaneous | 5 | 2000 | 10000 |