MIAI Smart Gesture Controlled Selfie Drone
Nowadays, Social Media is taking the world to a whole new standard of regularly updating your friends and fans, but it's hard to deal with such efforts when you are lonely enjoying on a trip and missing someone around you to take care of your photos/videos. The "selfie" is the word for a decade now
2025-06-28 16:28:34 - Adil Khan
MIAI Smart Gesture Controlled Selfie Drone
Project Area of Specialization Computer ScienceProject SummaryNowadays, Social Media is taking the world to a whole new standard of regularly updating your friends and fans, but it's hard to deal with such efforts when you are lonely enjoying on a trip and missing someone around you to take care of your photos/videos. The "selfie" is the word for a decade now where more and more people are taking it and posting it on all forms of Social Media Platforms. With selfie sticks making rounds now, it would be a good idea to use a Gesture Controlled Drone for this purpose. It's one of the most exciting uses of the drone. So no selfie sticks would hinder your creativity while you are on the go and no one around you to take your cinematic shots.
Artificial Intelligence stepped into this research of person and its pose detection using computer vision.
The software product which will be developed is based on the implementation of an AI Computer Vision Algorithm for normal users to smartly operate Complex Drone tasks with ease. In addition to manually operating the drone, we are providing Gesture-Based Controlling Feature along with Person Detection that keeps him in the center of video view in an open environment for Cinematic Videos and Photos
For the travel-vloggers, capturing the moment is their passion, and when they are alone and wandering at a location where there is no one to click photo or video of them, this is where it comes in and cinematically takes the profile shots. They need just to move around and do their fun while this drone gets them covered.
Project ObjectivesThe system will be designed to be accessible by a normal user to take the shots from a wider angle without getting into technicality. The algorithm will efficiently be designed that the drone can be operated by a normal user to take the shots from a much wider and broader perspective without getting into technicalities.
The modern era is about advancements in photography and videography. This drone will lead the young generation to it in a different way.
Objectives of the system are:
- To eliminate the requirement of having the expertise to operate the drone.
- Capture shots from unique perspectives that can never be achieved with selfies.
- Provide HD clear shots that mostly smartphone’s selfie camera lacks.
- Manually controlled with just a few clicks.
- No matter where you go in open surroundings at walking speed, it will follow.
- Safely land the drone.
- Control the Drone using gestures.
Since the system is semi hardware-based, it requires a Quadcopter to take the flight in the air. Along with that, a Raspberry PI 4 is needed to run synchronous AI and CV algorithms that would detect the person and Gesture Detection algorithm would check for any possible actions performed by the user as a gesture and those movement commands need to be sent to the drone using Wi-Fi while keeping its state in the memory.
Project Implementation MethodOur project is both hardware and software-based. We will develop an application for controlling a Drone. When firstly the user will run the Drone application, a message to connect to the drone’s Wi-Fi will appear. Along with AI Person Tracking Algorithm, Live Feed from Selfie Quadcopter will be streamed. User will be able to see video results receiving from the drone as well as control the drone. All this video transmission and control will be done on the standard Wi-Fi connection from the controlling device to the drone. Video feed from the drone will be processed using Neural Network frame by frame. It returns the coordinates of the detected person or the object. Further, Drone will move towards that person with a predefined or changeable distance at a defined movement of speed.
Video from the drone will be received in the form of a frame. Every frame the camera captures will be transmitted to the controller computer. Such a live-view feature will be extremely helpful for the user especially in Manual Control mode. Also, the synchronous recording will make video recordings more reliable (no data loss at a crash) and more practical (user can use it according to his/her desire from the very first moment with no waiting time).
As Drone is basically a gesture-controlled drone, so there are multiple gestures trained on a dataset. The image will be processed receiving from the drone using the model that tells what type of gesture is being performed. Respective actions will be taken regarding that gesture.
Benefits of the ProjectOur Smart Drone System will benefit the audience who would like to get updated on social media platforms and apps. A person is likely to be tracked using an advanced AI algorithm that predicts the location of the user and hence makes the Drone move. The Drone will move in such a direction where the user stands in the center of the view. So in this way, human mistakes while manually flying the drone can be eliminated.
From taking-off to landing the Drone, it's made autonomous to make it easy to fly in different open environments by an ordinary user without knowing the complex knowledge required to control the Pitch, Yaw, and Roll movements for the Drone.
As there are some laws applicable on the Drone by FPAA that a license is required for a Drone package weighing more than 250g, so for this purpose, we found a perfect Drone that only weighs 80g and doesn’t violate the law. So, it can be highly adaptive while traveling and can easily fit in a backpack.
It can fly around you even when there is no one to capture the moments you want to share with everyone or just get it to your gallery album.
Technical Details of Final DeliverableCommunication: The Drone will communicate with the Raspberry PI using an ordinary Wi-Fi connection and commands are sent using a UDP connection on a local IP and a unique port number. The program is written in Python language along with Artificial Intelligence and Computer Vision algorithms.
HD Camera: A HD 5MP with 720p wide resolution camera will transfer the images/videos which will be used for recording and processing by AI Algorithm to locate the person in the frame along with detection of any gesture.
State: Drone can fly in two modes which can be a Manual Flight or a Smart Mode Flight to move around the Drone in an environment.
Movement Commands: The user can take-off or land with a single click of a button, which results in AI maneuver that controls and sends appropriate commands to avoid any human error as it mostly happens due to its complexity. There must be a single choice for a user to do the flight operation. During the Smart Flight Mode, Drone automatically tries to detect a person. If a person happens in the frame, the Drone will calculate the direction and distance it needs to move. This will result in Drone following the person and detecting any possible gestures that a user is acquiring to have. Those gestures can be either used to take the photos and videos or they can be to do an Automatic Landing to avoid any user interaction of a Drone Complex Controller.
Final Deliverable of the Project HW/SW integrated systemCore Industry MediaOther IndustriesCore Technology RoboticsOther Technologies Artificial Intelligence(AI)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) | 79000 | |||
| Tello Drone Combo | Equipment | 1 | 39500 | 39500 |
| Raspberry Pi 4 (8GB) | Equipment | 1 | 19000 | 19000 |
| WiFi Module for Raspberry Pi | Equipment | 1 | 1500 | 1500 |
| Tello’s Additional Battery | Equipment | 2 | 4000 | 8000 |
| Tello’s Protection Cage | Equipment | 1 | 1000 | 1000 |
| Others | Miscellaneous | 1 | 10000 | 10000 |