Invisible Cloak using python and Opencv

3 min readSep 5, 2021

Hi, everyone. We’re trying to implement some magic today by creating code in our code editor. So you’ve heard about Harry Potter’s Invisible Cloak and want to get your hands on one so you can go wherever you want and experience what it’s like to be invisible.

Coder === Magician

About Invisible Cloak Project

We will create the invisible cloak using an image processing technique called Color detection and segmentation. In order to make this project, you’ll need a single-color cloth.

Note:

Cloth color should be unique relative to the background. (i.e. If the cloth is green then the background shouldn’t contain any green color)
In good lighting conditions, it works best.
Try to choose red, green or blue cloth, because these three colors are easier to detect.

we will use a red cloth for this project for Color detection is a technique where we can detect any color in a given range of HSV color space.

Lets try in the code

To install opencv, run the below command

!Pip install opencv-python

Step 1 — Import necessary packages and Initialize the camera

import cv2import numpy as npimport time

Step 2 — Find the available camera

#check for available cameraall_camera_idx_available = []for camera_idx in range(10):cap = cv2.VideoCapture(camera_idx,cv2.CAP_DSHOW)if cap.isOpened():print(f'Camera index available: {camera_idx}')all_camera_idx_available.append(camera_idx)cap.release()else:print("not available")

Step 3 - capture the background

# now capture the video  check for avialble camera devices as mentioned abovevideo = cv2.VideoCapture(0)print(video.isOpened())time.sleep(3) #timer to on the camfor i in range(60): #loopig to get a good image    rel,background = video.read() #read the background    if not rel:       raise RuntimeError("Couldn't capture an image")background = np.flip(background, axis=1) #filp the captured img

Step 4— Detect the cloth:

while(video.isOpened()): # untill the video is oprenrel,img=video.read()if rel==False:break #checking weather its workingimg=np.flip(img,axis=1)hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV) #convert to hsv detect colours betterlower_red = np.array([0,120,50])upper_red = np.array([10,255,255])mask1 = cv2.inRange(hsv, lower_red, upper_red)lower_red = np.array([170,120,70])  #range for red we wantupper_red = np.array([180,255,255])mask2 = cv2.inRange(hsv, lower_red, upper_red)mask1 = mask1+mask2 #detect all ranges of redmask1 = cv2.morphologyEx(mask1, cv2.MORPH_OPEN, np.ones((3,3), np.uint8),iterations=2) #removing the noice from datamask1 = cv2.morphologyEx(mask1, cv2.MORPH_DILATE, np.ones((3,3), np.uint8))mask2 = cv2.bitwise_not(mask1)res1 = cv2.bitwise_and(img,img, mask=mask2) #person who holds the cloakres2 = cv2.bitwise_and(background,background, mask=mask1)#to replace the background image in place of the colour red combining resultsfinal = cv2.addWeighted(res1, 1, res2, 1, 0)cv2.imshow("final",final)key = cv2.waitKey(1)#to close the windowif key== ord('c'):breakvideo.release()cv2.destroyAllWindows()

why we used hsv?

HSV stands for HUE, SATURATION, and VALUE (or brightness). It is a cylindrical color space

applying mask :

res1 = cv2.bitwise_and(img,img, mask=mask2) #person who holds the cloakres2 = cv2.bitwise_and(background,background, mask=mask1)

combining the mask :

#to replace the background image in place of the colour red combining resultsfinal = cv2.addWeighted(res1, 1, res2, 1, 0)

GitHub - baasithshiyam/invisible-cloak

Contribute to baasithshiyam/invisible-cloak development by creating an account on GitHub.

github.com

https://youtu.be/yAK0wLlvPHM

We implemented color detection and segmentation technique. In this opencv project, we learned about morphological operations, masking, and other image processing

video Demonstration

https://youtu.be/yAK0wLlvPHM

Done