improved_digit_detection.py

import cv2
import mediapipe as mp
from pprint import pprint
from easy_vector import Vector
from functools import reduce
from time import sleep

# https://pypi.org/project/vectors/

mp_drawing = mp.solutions.drawing_utils
mp_hands = mp.solutions.hands
landmark = mp_hands.HandLandmark

def digitExtended(*args, digit="unidentified", extended_value_ratio=0.3):
    """Calculates whether or not the digits of a finger are in a straight line."""
    if len(args) < 2:
        return 0
    # First, normalize by subtracting the first vector from all other vectors.
    args2 = list(map(lambda v: v - args[0], args))

    # Subtract the accumulated previous points from each point to isolate each vector.
    points = []
    acc = Vector(0,0,0)
    for index in range(len(args2)):
        vec = args2[index] - acc
        points.append(vec)
        acc = acc + args2[index]
        # print(f"Point {index} : {vec}")
        

    #pprint(points)
    
    # Next, add all the vectors together to create a final vector
    length = 0
    vecTotal = Vector(0,0,0)
    for point in points:
        length = length + point.length

    vecTotal = args[len(args)-1] - args[0]

    #pprint(length)
    #pprint(vecTotal)
    difference = vecTotal.length/length
    #pprint(difference)
    return (digit, difference > extended_value_ratio, length, vecTotal.length, difference)

# For webcam input:
cap = cv2.VideoCapture(0)
with mp_hands.Hands(
    min_detection_confidence=0.6,
    min_tracking_confidence=0.6) as hands:
  while cap.isOpened():
    success, image = cap.read()
    if not success:
      print("Ignoring empty camera frame.")
      # If loading a video, use 'break' instead of 'continue'.
      continue

    # Flip the image horizontally for a later selfie-view display, and convert
    # the BGR image to RGB.
    image = cv2.cvtColor(cv2.flip(image, 1), cv2.COLOR_BGR2RGB)
    # To improve performance, optionally mark the image as not writeable to
    # pass by reference.
    image.flags.writeable = False
    results = hands.process(image)

    # Draw the hand annotations on the image.
    image.flags.writeable = True
    image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
    if results.multi_hand_landmarks:
      for hand_landmarks in results.multi_hand_landmarks:
        mp_drawing.draw_landmarks(image, hand_landmarks, mp_hands.HAND_CONNECTIONS)
        points = list(map(lambda v: Vector(v.x, v.y, v.z), hand_landmarks.landmark))
        # Fingers are 0-thumb 1-index, 2- middle, etc. Creates boolean list of 'extended' or not.
        fingers_extended = [
            digitExtended(points[1], points[2], points[3], points[4], digit="thumb ", extended_value_ratio=1.0),
            digitExtended(points[0], points[5], points[6], points[7], points[8], digit="index "),
            digitExtended(points[0], points[9], points[10], points[11], points[12], digit="middle"),
            digitExtended(points[0], points[13], points[14], points[15], points[16], digit="ring  "),
            digitExtended(points[0], points[17], points[18], points[19], points[20], digit="pinkie"),
        ]

        #pprint(fingers_extended)
        debug_str = "\n\n============================"
        for finger in fingers_extended:
            debug_str = debug_str + f"\n{finger[0]} => {finger[1]}  :   len: {finger[2]:.2f}   vecLen: {finger[3]:.2f}   diff: {finger[4]:.2f}"
        print(debug_str)
        #sleep(0.2)

    cv2.imshow('MediaPipe Hands', image)
    if cv2.waitKey(5) & 0xFF == 27:
      break
cap.release()