Intelligent surveillance system
We are living in a world where everything is revolutionizing either if we are doing a simpler task as to study from home or the more complex one as to explore the galaxies, everything is bringing revolution and ease in it of itself. Similarly, the ways to maintain security is also revolutionizing. I
2025-06-28 16:33:20 - Adil Khan
Intelligent surveillance system
Project Area of Specialization Artificial IntelligenceProject SummaryWe are living in a world where everything is revolutionizing either if we are doing a simpler task as to study from home or the more complex one as to explore the galaxies, everything is bringing revolution and ease in it of itself. Similarly, the ways to maintain security is also revolutionizing. In the past we use to keep a dog at our door to detect the coming of an unknown person. Then the world enhanced a bit, and we made surveillance camera whose live stream video was constantly viewed by a person on a monitor screen and still it is in practice. Then everything started to become intelligent so as the surveillance cameras, without any human intervention, start to detect motion. Firstly, of anything and then of humans. The motion detection either works by using infrared radiation sensors or by advanced techniques like using computer vision algorithms. Still there was a problem of numerous notifications of motion sensing. A lot of which was based on the motion of a known person rather than a suspicious one. Thus, we came up with an idea to use deep learning techniques to train our model using computer vision algorithms and a large dataset of surveillance videos. So that it cannot only distinguish between a pet or a human motion but can also distinguish motion as a normal activity or a suspicious activity and can generate alarms only when the movement is suspected as unfavorable.
Project ObjectivesOur first and main objective is to help secure an environment using the latest technology i.e. Neural Networks and deep learning.
Our second objective is to bring ease in the lives of people by making them feel secure.
Project Implementation Method- The first step is to train the model to recognize object and label them as dog, human, car, truck etc. using CNN model.
- The second step is to train the model to detect motion of human
- The third main objective is to train the model to detect action and categorized those action into different categories using deep learning model e.g. running, walking etc.
- Then to generate alarm when the activity category in undesirable as well as to start recording automatically so that the user can view what has happened and why was the alarm generated.
- We need to connect the camera to our android mobile app.
- All the functionality will be done using the attached RAM to the raspberry pi camera to save the battery of the mobile. Hence in the mobile only the alarms will be generated plus the live stream video and the recorded videos can be viewed.
- Also, we will provide some functionality in the android app. Like the ability to record videos on a certain time and date.
Clearly, while offering the essential features, today’s video surveillance system fails to provide the level of detail required to shift the security paradigm from “investigation to preemption”. Technologies for automated visual analysis will shift the video surveillance system of today from investigation to the preventive paradigm. Intelligent Surveillance Systems have a range of advantages over conventional video surveillance systems, including.
- the ability to avoid incidents through real-time suspicious activity warnings.
- enhanced forensic capability through content-based video retrieval.
- situation awareness by mutual awareness of the location and behavior of objects in the space being monitored.
Final deliverable will be a system including an android app and a surveillance camera along with RAM and hard-drive for running of the system.
Surveillance camera will be of rasberry pie. RAM and GPU card will also be needed. Heavy ROM for long term video storage will also be one of our deliveravle.
Final Deliverable of the Project Software SystemCore Industry SecurityOther Industries IT Core Technology Internet of Things (IoT)Other Technologies Artificial Intelligence(AI), NeuroTechSustainable Development Goals Decent Work and Economic Growth, Peace and Justice Strong InstitutionsRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 69743 | |||
| 90 Degree Angle Bracket | Equipment | 1 | 4000 | 4000 |
| USB Micro-B Cable - 6 | Equipment | 1 | 313 | 313 |
| #8 Wing Nut | Equipment | 1 | 1765 | 1765 |
| #8 Hex Nut | Equipment | 1 | 160 | 160 |
| #8 X 3/4 | Equipment | 1 | 160 | 160 |
| M2.5 Machine Screws | Equipment | 1 | 160 | 160 |
| M4 Hex Nuts | Equipment | 1 | 200 | 200 |
| M4 X 20mm Long Machine Screws | Equipment | 1 | 200 | 200 |
| Outdoor Mounting Tape | Equipment | 1 | 800 | 800 |
| 1/4 | Equipment | 1 | 2208 | 2208 |
| Micro-USB t USB Cable (Generic) | Equipment | 1 | 1115 | 1115 |
| Adafruit Raspberry Pi Zero Camera Cable | Equipment | 1 | 955 | 955 |
| Raspberry Pi Camera Module | Equipment | 1 | 3871 | 3871 |
| Raspberry Pi Zero Wireless | Equipment | 1 | 5836 | 5836 |
| Graphic processing unit (GPU)(GEFORCE RTX 3090) | Equipment | 1 | 48000 | 48000 |