Open In App

Live Webcam Drawing using OpenCV

Last Updated : 11 Aug, 2025
Comments
Improve
Suggest changes
4 Likes
Like
Report

Live webcam drawing allows you to track and draw the movements of colored objects in real time. Using OpenCV, we capture video from the webcam, detect specified colors and create contours around the detected objects. These movements are then painted on a virtual canvas, giving the effect of "drawing in the air."

Prerequisites:

Install them using pip if not already installed:

pip install opencv-python numpy

Approach

  1. Capture video from the webcam and set frame size and brightness.
  2. Define HSV color ranges for objects to track.
  3. Convert frames to HSV format and create masks for each color.
  4. Find contours to locate the center of detected objects.
  5. Store points and draw them on a virtual canvas.
  6. Continue updating the canvas in real time until the user quits.

Python Implementation

Python 
import cv2
import numpy as np

W, H = 640, 480  # Frame width & height

cap = cv2.VideoCapture(0)
cap.set(3, W)
cap.set(4, H)
cap.set(10, 150)

# HSV ranges for tracking colors
color_ranges = [
    [5, 107, 0, 19, 255, 255],
    [133, 56, 0, 159, 156, 255],
    [57, 76, 0, 100, 255, 255],
    [90, 48, 0, 118, 255, 255]
]

# Corresponding BGR values for drawing
draw_colors = [
    [51, 153, 255],
    [255, 0, 255],
    [0, 255, 0],
    [255, 0, 0]
]

points = []  # [x, y, color_id]


def find_color(img, ranges, bgr_vals):
    hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
    new_pts = []
    for idx, rng in enumerate(ranges):
        lower, upper = np.array(rng[:3]), np.array(rng[3:])
        mask = cv2.inRange(hsv, lower, upper)
        cx, cy = get_contour_center(mask)
        cv2.circle(result, (cx, cy), 15, bgr_vals[idx], cv2.FILLED)
        if cx != 0 and cy != 0:
            new_pts.append([cx, cy, idx])
    return new_pts


def get_contour_center(mask):
    contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
    cx, cy, w, h = 0, 0, 0, 0
    for cnt in contours:
        if cv2.contourArea(cnt) > 500:
            peri = cv2.arcLength(cnt, True)
            approx = cv2.approxPolyDP(cnt, 0.02 * peri, True)
            x, y, w, h = cv2.boundingRect(approx)
            return x + w // 2, y
    return cx, cy


def draw_canvas(pts, bgr_vals):
    for x, y, color_id in pts:
        cv2.circle(result, (x, y), 10, bgr_vals[color_id], cv2.FILLED)


while True:
    ret, frame = cap.read()
    result = frame.copy()
    new_pts = find_color(frame, color_ranges, draw_colors)
    if new_pts:
        points.extend(new_pts)
    if points:
        draw_canvas(points, draw_colors)

    cv2.imshow("Result", result)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

Output : 

Explanation:

  • cap.set(3, W) & cap.set(4, H) set webcam frame size and cap.set(10, 150) adjust brightness.
  • cv2.inRange() creates mask for isolating the selected color range.
  • get_contour_center() finds object’s center from contours in the mask.
  • find_color() detects objects, marks their positions and returns coordinates with color IDs.
  • draw_canvas() draws stored points on the live feed, shown via cv2.imshow().

R

romilvishol
Improve
Article Tags :

Explore

Lightbox