ASME Software Task-2: Shuri's Gesture Vault

Continuing the saga from Daredevil's Echo Hunt — now, a new hero rises in Wakanda!

What This Project Does

two main applications that use computer vision to recognize hand gestures:

Q1: Gesture Recognition Engine

A gesture-controlled vault system that responds to hand movements:

• Open Palm → Unlocks the vault
• Closed Fist → Locks the vault
• Swipe Left/Right → Navigate security menus
• Bonus Features: Gesture passwords and custom gesture recording

Q2: Virtual Mouse 🖱

Control your computer mouse using just your hand:

• Open Palm → Move cursor around
• Pinch (Thumb + Index) → Left click
• Closed Fist → Right click
• Index Finger Only → Freeze cursor

Tech Stack

• Python - Main programming language
• OpenCV - For camera feed and image processing
• MediaPipe - Google's hand tracking library (this thing is amazing!)
• PyAutoGUI - For controlling mouse movements
• NumPy - Mathematical operations

How to Use

Gesture Vault Controls:

• Show an open palm to unlock
• Make a fist to lock
• Swipe left/right with your index finger to navigate
• Try the secret gesture passwords: Palm → Fist → Palm → Swipe Right

Virtual Mouse Controls:

• Open your hand to move the cursor
• Pinch thumb and index finger to left-click
• Close all fingers to right-click
• Point with index finger to freeze cursor

Cool Features I Added

• Real-time gesture logging with timestamps
• Gesture password sequences for extra security
• Custom gesture recording - you can teach it new gestures
• Smooth cursor movement with jitter reduction
• Visual feedback showing current gesture and actions

Challenges I Faced

1. Camera permissions on macOS - took forever to figure out
2. Gesture detection accuracy - thumb detection was tricky
3. Lighting conditions - works best in good lighting