CVZone

This is a Computer vision package that makes its easy to run Image processing and AI functions. At the core it uses OpenCV and Mediapipe libraries.

Installation

You can simply use pip to install the latest version of cvzone.

pip install cvzone

---

60 FPS Face Detection

---

from cvzone.FaceDetectionModule import FaceDetector
import cv2

cap = cv2.VideoCapture(0)
detector = FaceDetector()

while True:
    success, img = cap.read()
    img, bboxs = detector.findFaces(img)

    if bboxs:
        # bboxInfo - "id","bbox","score","center"
        center = bboxs[0]["center"]
        cv2.circle(img, center, 5, (255, 0, 255), cv2.FILLED)

    cv2.imshow("Image", img)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

---

Hand Tracking

---

Basic Code Example

from cvzone.HandTrackingModule import HandDetector
import cv2

cap = cv2.VideoCapture(0)
detector = HandDetector(detectionCon=0.8, maxHands=2)
while True:
    # Get image frame
    success, img = cap.read()
    # Find the hand and its landmarks
    hands, img = detector.findHands(img)  # with draw
    # hands = detector.findHands(img, draw=False)  # without draw

    if hands:
        # Hand 1
        hand1 = hands[0]
        lmList1 = hand1["lmList"]  # List of 21 Landmark points
        bbox1 = hand1["bbox"]  # Bounding box info x,y,w,h
        centerPoint1 = hand1['center']  # center of the hand cx,cy
        handType1 = hand1["type"]  # Handtype Left or Right

        fingers1 = detector.fingersUp(hand1)

        if len(hands) == 2:
            # Hand 2
            hand2 = hands[1]
            lmList2 = hand2["lmList"]  # List of 21 Landmark points
            bbox2 = hand2["bbox"]  # Bounding box info x,y,w,h
            centerPoint2 = hand2['center']  # center of the hand cx,cy
            handType2 = hand2["type"]  # Hand Type "Left" or "Right"

            fingers2 = detector.fingersUp(hand2)

            # Find Distance between two Landmarks. Could be same hand or different hands
            length, info, img = detector.findDistance(lmList1[8], lmList2[8], img)  # with draw
            # length, info = detector.findDistance(lmList1[8], lmList2[8])  # with draw
    # Display
    cv2.imshow("Image", img)
    cv2.waitKey(1)
cap.release()
cv2.destroyAllWindows()

---

Pose Estimation

---

from cvzone.PoseModule import PoseDetector
import cv2

cap = cv2.VideoCapture(0)
detector = PoseDetector()
while True:
    success, img = cap.read()
    img = detector.findPose(img)
    lmList, bboxInfo = detector.findPosition(img, bboxWithHands=False)
    if bboxInfo:
        center = bboxInfo["center"]
        cv2.circle(img, center, 5, (255, 0, 255), cv2.FILLED)

    cv2.imshow("Image", img)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

---

Face Mesh Detection

---

from cvzone.FaceMeshModule import FaceMeshDetector
import cv2

cap = cv2.VideoCapture(0)
detector = FaceMeshDetector(maxFaces=2)
while True:
    success, img = cap.read()
    img, faces = detector.findFaceMesh(img)
    if faces:
        print(faces[0])
    cv2.imshow("Image", img)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

---

Stack Images

---

import cvzone
import cv2

cap = cv2.VideoCapture(0)
cap.set(3, 1280)
cap.set(4, 720)

while True:
    success, img = cap.read()
    imgGray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    imgList = [img, img, imgGray, img, imgGray, img,imgGray, img, img]
    stackedImg = cvzone.stackImages(imgList, 3, 0.4)

    cv2.imshow("stackedImg", stackedImg)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

---

Corner Rectangle

---

---

import cvzone
from cvzone.HandTrackingModule import HandDetector
import cv2

cap = cv2.VideoCapture(0)
detector = HandDetector()

while True:
    # Get image frame
    success, img = cap.read()

    # Find the hand and its landmarks
    img = detector.findHands(img, draw=False)
    lmList, bbox = detector.findPosition(img, draw=False)
    if bbox:
        # Draw  Corner Rectangle
        cvzone.cornerRect(img, bbox)

    # Display
    cv2.imshow("Image", img)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

FPS

---

import cvzone
import cv2

fpsReader = cvzone.FPS()
cap = cv2.VideoCapture(0)
cap.set(3, 1280)
cap.set(4, 720)

while True:
    success, img = cap.read()
    fps, img = fpsReader.update(img,pos=(50,80),color=(0,255,0),scale=5,thickness=5)
    cv2.imshow("Image", img)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

Fingers UP

---

The cvzone library uses the fingersUp function of the HandDetector class of the HandTrackingModule module to count open fingers. This function gives the correct answer for the normal state where the hand is shown vertically to the camera, but if you tilt the hand, turn the hand upside down or hold the back of the hand to the camera, you cannot get a correct answer from this function. As an example, you can see the video below.

In this version, I have upgraded this function to cover the previously mentioned cases (tilting the hand, turning the hand upside down, showing the back of the hand). As an example, you can see the video below.

To do the mentioned things, we first measure the angle of the hand with the vertical axis and then rotate the LandMarkList by the same amount to send the correct information to the main function.