Drone Based Surveillance Using Real Time Image Processing For Detection Of Illegal Activity
The project is based on Surveillance through drones using image processing. We provide a solid solution in our proposal. The drone is a "robotic" autonomous aircraft, which use in various fields like military, media, emergency rescue, outer space, and event coverage. The crime took place mostly wher
2025-06-28 16:26:52 - Adil Khan
Drone Based Surveillance Using Real Time Image Processing For Detection Of Illegal Activity
Project Area of Specialization Computer ScienceProject SummaryThe project is based on Surveillance through drones using image processing. We provide a solid solution in our proposal. The drone is a "robotic" autonomous aircraft, which use in various fields like military, media, emergency rescue, outer space, and event coverage. The crime took place mostly where the security system is low or police are not rapidly accessing the area. Such as crimes in complex environments. Now a day’s different security techniques are being used such as CCTV, Metal detection, etc. But crimes also be happened due to different reasons like points not covered by CCTV, real-time surveillance through drones is a good solution. We use a different algorithms to detect illegal activity in surrounding using image processing. We used ssd mobilenet v2 as a shape detecting algorithm to detect weapons used for crimes, SVM is a classification algorithm, and detection is taking place by image training. The main goal is to maintain peace.
Project ObjectivesThe main objective to build this project to provide remote surveillance by a drone that is capable of real-time image processing for detecting illegal activity. The whole project is a solution of self-monitoring and self-targeting with a manual control system that uses of real-time object detection techniques of a targeted environment by AI UAV security surveillance drone that eliminates expenses, budgets, and human efforts to cover all crime scenes and weapons detection in an effective manner.
Project Implementation MethodIn this project, we divided our project into two phases. the first phase of the project is based on making quadcopter drones and the second phase is based on software making for the detection of illegal activity. for detecting the custom object in an image we used a histogram of the oriented gradient algorithm in our project. first, we assembled the drone then we programed it onto mission planner software. after that we started working on software we trained our custom model with help of TensorFlow. We collected the images data from the internet and label it for training. we used Google Colab GPU for our custom model training. after the training, we embedded our custom model with the main detection code. Then we deployed it onto raspberry pi. After that, we integrated the drone with software and hardware for a final product of our model.
Benefits of the ProjectThe project has great benefits for security agencies to conduct remote surveillance in places where they can't put CCTV cameras or security officers due to a lack of resources. This project presents AI base security surveillance UAV Drone for detecting illegal activities in specific areas to provide. . The whole project is a solution of self-monitoring and self-targeting with a manual control system that uses real-time object detection techniques of a targeted environment by AI UAV security surveillance drone that eliminates expenses, budgets, and human efforts to cover all crime scenes and weapons detection in an effective manner. The drone can be beneficial for events as well there we cannot deploy forces for surveillance.
Technical Details of Final DeliverableThe technical details of the project are as follows:
1) The drone as the mainboard is APM 2.8 flight controller. APM 2.8 is the brain of a drone that has a drone as operating software. It is based on Arduino Mega 2560 programmable circuit board that contains different sensors such as Gyroscope, Accelerometer, Barometer, and Magnetometer. All the sensors are used to control the drone.
2) Six-channel receiver is used in the project that is connected with the flight controller and this receiver collects controlling commands from the ground controlling station transmitter.
3) Raspberry pi is used as for performing detection task that contains developed program on it.
4) Raspberry pi IR camera module is used for taking input from the environment.
5) The input is the form of live footage after the input the final task is analyzing the footage if it contains any person who is performing snatching or robbery it will capture the image and then send it with SMTP Google server to the concerned authorities
Final Deliverable of the Project HW/SW integrated systemCore Industry SecurityOther Industries IT Core Technology Internet of Things (IoT)Other Technologies OthersSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 70000 | |||
| DJI Flamewheel F450 Frame | Equipment | 1 | 2000 | 2000 |
| ESC Electronic Speed Control | Equipment | 4 | 850 | 3400 |
| Brushless BLDC Motors | Equipment | 4 | 850 | 3400 |
| Propellers 10x45 Pair | Equipment | 2 | 250 | 500 |
| Landing Gears | Equipment | 1 | 450 | 450 |
| APM 2.8 | Equipment | 1 | 5700 | 5700 |
| FlySky 6 Channel Receiver | Equipment | 1 | 2600 | 2600 |
| FlySky FS-I6AB Transmitter | Equipment | 1 | 9500 | 9500 |
| Sik Telemetry Radio 433mhz | Equipment | 1 | 4800 | 4800 |
| M8N GPS Module | Equipment | 1 | 5400 | 5400 |
| APM Power Module | Equipment | 1 | 950 | 950 |
| LIPO Battery 12.3 Volts | Equipment | 1 | 4600 | 4600 |
| Raspberry Pi 4 2GB | Equipment | 1 | 15000 | 15000 |
| Raspberry Pi IR Night Vision Camera Module | Equipment | 1 | 4500 | 4500 |
| Micro SD card 32GB class 10 | Equipment | 1 | 1000 | 1000 |
| Voltage Dropper 12v to 5v | Equipment | 1 | 200 | 200 |
| others | Miscellaneous | 1 | 6000 | 6000 |