Project Summary – Gesture-Controlled Asphalt 8: Airbone
This Python program allows the game Asphalt 8 to be controlled using hand gestures instead
of a keyboard. It uses computer vision and AI-based hand tracking for real-time gesture
recognition.
⚙ How It Works:
1. Libraries Used:
o OpenCV (cv2): Captures and displays live video from the webcam.
o MediaPipe: Detects and tracks hand landmarks (fingers, joints, and palm).
o PyAutoGUI / pyDirectInput: Simulates keyboard presses for in-game control.
2. Hand Detection:
The webcam feed is processed through MediaPipe Hands, which identifies each hand.
3. Gesture Recognition:
Based on the finger positions, the program identifies three main gestures:
o Fist – used for turning left right or braking
o Palm – used for acceleration
o Victory – used for jumping (space key)
4. Game Control Mapping:
o Both Palms → Press “W” for acceleration
o Left Fist + Right Palm → Press “A” for left turn
o Right Fist + Left Palm → Press “D” for right turn
o Both Fists → Press “S” for brake
o Any Victory Sign → Press “Space” for jump
⚙ In Short
This code combines AI hand tracking with keyboard automation to create a gesture-based
control system for gaming — turning human hand gestures directly into game actions.
Project Summary – Gesture-Controlled Google Dino
This Python program lets you play the Google Chrome Dino game using hand gestures
instead of the keyboard. It uses AI-based hand tracking to detect when a player makes a fist
and automatically makes the Dino jump.
⚙ How It Works:
1. Libraries Used:
o OpenCV (cv2): Captures real-time video from the webcam.
o MediaPipe: Detects and tracks the hand and its finger landmarks using machine
learning.
o PyAutoGUI: Simulates keyboard presses to control the game (pressing the
spacebar).
2. Hand Detection:
The webcam feed is processed through MediaPipe Hands, which detects one hand.
3. Gesture Recognition:
A custom function is_fist() checks if all fingers are folded by comparing the position
of fingertips and their joints.
4. Game Control Mapping:
o Fist → Presses “Space” key to make the Dino jump
o Open Hand → No action, Dino keeps running straight
⚙ In Short
The program combines computer vision (OpenCV) and AI gesture tracking (MediaPipe) to
let players control the Google Dino game with just hand movements, making gameplay touch-
free, interactive, and fun.
Project Summary – Gesture-Controlled Hill Climb Racing
This Python program allows players to control Hill Climb Racing using hand gestures instead
of a physical keyboard or controller. It uses AI-based hand tracking and a virtual gamepad
interface to simulate real-time driving actions like acceleration and braking.
⚙ How It Works:
1. Libraries Used:
o OpenCV (cv2): Captures and displays live video from the webcam.
o MediaPipe: Detects and tracks both hands using machine learning-based hand
landmarks.
o vgamepad: Creates a virtual Xbox controller for smooth, real-time game
control.
o PyAutoGUI: Used for pressing other keyboard keys (like Enter) when needed.
2. Hand Detection:
MediaPipe detects two hands.
Each frame is analyzed to recognize if the hand is a fist or a palm using landmark
position comparison.
3. Gesture Recognition:
o Fist: All fingers folded — represents a ―press‖ action.
o Palm: All fingers extended — represents an ―idle‖ or ―release‖ action.
1. Game Control Mapping:
o Right Fist + Left Palm → Move Right
o Left Fist + Right Palm → Move Left
o Both Fists → Move Down
o Both Palms → No Action (neutral state)
o Any Victory Sign → Press “W” for jump
⚙ In Short
This project combines MediaPipe’s AI gesture tracking, OpenCV video processing, and
vgamepad virtual controller technology to create a hands-free, gesture-controlled gaming
experience — making Hill Climb Racing more interactive and futuristic.
Project Summary – Gesture-Controlled Temple Run
This Python program allows players to control Temple Run using hand gestures instead of a
physical keyboard or controller. It uses AI-based hand tracking and a virtual gamepad
interface to simulate real-time running actions like jumping, sliding, and turning.
⚙ How It Works:
2. Libraries Used:
o OpenCV (cv2): Captures and displays live video from the webcam.
o MediaPipe: Detects and tracks both hands using machine learning-based hand
landmarks.
o vgamepad: Creates a virtual Xbox controller for smooth, real-time game
control.
o PyAutoGUI: Used for pressing other keyboard keys (like Enter) when needed.
3. Hand Detection:
MediaPipe detects two hands.
Each frame is analyzed to recognize if the hand is a fist or a palm using landmark
position comparison.
4. Gesture Recognition:
o Fist: All fingers folded — represents a ―press‖ action.
o Palm: All fingers extended — represents an ―idle‖ or ―release‖ action.
5. Game Control Mapping:
o Right Fist + Left Palm → Move Right
o Left Fist + Right Palm → Move Left
o Both Fists → Move Down
o Both Palms → No Action (neutral state)
o Any Victory Sign → Press “W” for jump
⚙ In Short
This project combines MediaPipe’s AI gesture tracking, OpenCV video processing, and
vgamepad virtual controller technology to create a hands-free, gesture-controlled gaming
experience — making Temple Run more interactive and futuristic.
Project Summary – Gesture-Controlled Subways Surfers
This Python program allows players to control Subways Surfers using hand gestures instead
of a physical keyboard or controller. It uses AI-based hand tracking and a virtual gamepad
interface to simulate real-time running actions like jumping, sliding, and turning.
⚙ How It Works:
1. Libraries Used:
o OpenCV (cv2): Captures and displays live video from the webcam.
o MediaPipe: Detects and tracks both hands using machine learning-based hand
landmarks.
o vgamepad: Creates a virtual Xbox controller for smooth, real-time game
control.
o PyAutoGUI: Used for pressing other keyboard keys (like Enter) when needed.
2. Hand Detection:
MediaPipe detects two hands.
Each frame is analyzed to recognize if the hand is a fist or a palm using landmark
position comparison.
3. Gesture Recognition:
o Fist: All fingers folded — represents a ―press‖ action.
o Palm: All fingers extended — represents an ―idle‖ or ―release‖ action.
4. Game Control Mapping:
o Right Fist + Left Palm → Move Right
o Left Fist + Right Palm → Move Left
o Both Fists → Move Down
o Both Palms → No Action (neutral state)
o Any Victory Sign → Press “W” for jump
⚙ In Short
This project combines MediaPipe’s AI gesture tracking, OpenCV video processing, and vgamepad
virtual controller technology to create a hands-free, gesture-controlled gaming experience — making
Subways Surfers more interactive and futuristic.