Creating a Finger Counter Using Computer Vision and OpenCv in Python
Last Updated :
03 Jan, 2023
Improve
In this article, we are going to create a finger counter using Computer Vision, OpenCv in Python
Required Libraries:
- OpenCV: OpenCV is a widely used library for image processing.
- cvzone: It is a computer vision package that makes it easy to run Image processing and AI functions
To install the Cvzone, run this command into terminal:
pip install cvzone
Stepwise Implementation:
Step 1: Import required libraries.
Python3
# import libraries and required classes import cv2 from cvzone.HandTrackingModule import HandDetector |
Step 2: Declare the HandDetector and open the default camera to capture the hand.
Python3
# declaring HandDetector with # some basic requirements detector = HandDetector(maxHands=1, detectionCon=0.8) # it detect only one hand from # video with 0.8 detection confidence video = cv2.VideoCapture(0) |
Step 3: Count the number of fingers
Capture the frames continuously and detect the hand from the frame then detect how many fingers are up and count them. On the counter make appropriate conditions and place an image of fingers.
Python3
while True: # Capture frame-by-frame _, img = video.read() img = cv2.flip(img, 1) # Find the hand with help of detector hand = detector.findHands(img, draw=False) # Here we take img by default if no hand found fing = cv2.imread("Put image path with 0 fingures up") if hand: # Taking the landmarks of hand lmlist = hand[0] if lmlist: # Find how many fingers are up # This function return list fingerup = detector.fingersUp(lmlist) # Change image based on # different-different conditions if fingerup == [0, 1, 0, 0, 0]: fing = cv2.imread("Put image path of 1\ fingures up") if fingerup == [0, 1, 1, 0, 0]: fing = cv2.imread("Put image path of 2\ fingures up") if fingerup == [0, 1, 1, 1, 0]: fing = cv2.imread("Put image path of\ 3 fingures up") if fingerup == [0, 1, 1, 1, 1]: fing = cv2.imread("sPut image path \ of 4 fingures up") if fingerup == [1, 1, 1, 1, 1]: fing = cv2.imread("Put image path \ of 4 fingures and thumbs up") # Resize the image fing = cv2.resize(fing, (220, 280)) # Place the image in main frame img[50:330, 20:240] = fing # Display the resulting frame cv2.imshow("Video", img) |
How actually detector detects the fingers up or down.
- It returns the list of five elements and each element depends on the finger’s condition.
- The list order like [ thumb, index finger, middle finger, ring finger, pinky/little finger ]
- If any of the fingers are up, it returns 1 for that particular index value else returns 0.
- And after making the list it will return by the function.
Step 4: Terminate the loop
Python3
# Enter key 'q' to break the loop if cv2.waitKey(1) & 0xFF == ord('q'): break # When everything done, release # the capture and destroy the windows video.release() cv2.destroyAllWindows() |
Below is the complete implementation:
Python3
import cv2 from cvzone.HandTrackingModule import HandDetector detector = HandDetector(maxHands=1, detectionCon=0.8) video = cv2.VideoCapture(1) while True: _, img = video.read() img = cv2.flip(img, 1) hand = detector.findHands(img, draw=False) fing = cv2.imread("Put image path with 0 fingures up") if hand: lmlist = hand[0] if lmlist: fingerup = detector.fingersUp(lmlist) if fingerup == [0, 1, 0, 0, 0]: fing = cv2.imread("Put image \ path of 1 fingures up") if fingerup == [0, 1, 1, 0, 0]: fing = cv2.imread("Put image \ path of 2 fingures up") if fingerup == [0, 1, 1, 1, 0]: fing = cv2.imread("Put image \ path of 3 fingures up") if fingerup == [0, 1, 1, 1, 1]: fing = cv2.imread("Put image \ path of 4 fingures up") if fingerup == [1, 1, 1, 1, 1]: fing = cv2.imread("Put image \ path of 4 fingures and thumbs up") fing = cv2.resize(fing, (220, 280)) img[50:330, 20:240] = fing cv2.imshow("Video", img) if cv2.waitKey(1) & 0xFF == ord('q'): break video.release() cv2.destroyAllWindows() |
Output :
