Hand Gesture Recognition using ESP32 and Python

The Hand Gesture Recognition system bridges computer vision software running on a host computer with hardware actuation controlled by an ESP32 microcontroller over a Wi-Fi network. The core operation begins with a standard webcam capturing real-time video frames of the user's hand. A Python script processes these frames using OpenCV for image handling and MediaPipe to detect hand landmarks and pinpoint the exact coordinates of each finger. By evaluating the relative positions of the finger joints, the script determines whether each individual finger (thumb, index, middle, ring, or pinky) is extended or folded. When a gesture state changes, the Python Requests library sends an HTTP request over the local wireless network to an asynchronous web server running directly on the ESP32 DevKit v1. The ESPAsyncWebServer library receives these network commands and updates the logic states of the microcontroller's digital output pins. Each GPIO pin is wired to a 220-ohm current-limiting resistor connected in series with an LED, ensuring that safe operating currents are maintained as the LEDs turn on or off in synchronization with the user's physical hand gestures.