Clone a NumPy array multiple times

Question

I loaded a picture into a numpy array and need to threshold it picture at 2 different thresholds.

import numpy as np
import cv2

cap = cv2.Videocapture(0)
_,pic = cap.read()
pic1 = pic
pic2 = pic

pic1[pic1 > 100] = 255
pic2[pic2 > 200] = 255

This code will always edit pic when I only want them to modify pic1 and pic2

Similar issue 4 hrs ago. Code actualising automatically variables when not desired - duplicate. When do different variables reference the same thing, as opposed to making a copy? — hpaulj
– hpaulj, Commented May 14, 2018 at 5:55

cs95 · Accepted Answer · 2018-05-14 06:12:42Z

4

In python, there is a difference between an object and a variable. A variable is name assigned to an object; and an object can have more than one name in memory.

By doing pic1 = pic; pic2 = pic, You're assigning the same object to multiple different variable names, so you end up modifying the same object.

What you want is to create copies using np.ndarray.copy—

pic1 = pic.copy()
pic2 = pic.copy()

Or, quite similarly, using np.copy—

pic1, pic2 = map(np.copy, (pic, pic))

This syntax actually makes it really easy to clone pic as many times as you like:

pic1, pic2, ... picN = map(np.copy, [pic] * N)

Where N is the number of copies you want to create.

edited May 14, 2018 at 6:12

answered May 14, 2018 at 5:55

cs95

406k106 gold badges744 silver badges797 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

cs95 Over a year ago

@liliscent To me, it's neater than pic1 = pic.copy(); pic2 = pic.copy(), though I suppose it's a matter of taste.

llllllllll Over a year ago

I somehow find this kind of simplification will make code less readable. But anyway, it's just personal preference.

cs95 Over a year ago

@liliscent Sure, I can respect that. I picked this up from Andras Deak, and found I quite liked the succinctness of it... of course it isn't for the faint of heart :p

Hojo.Timberwolf Over a year ago

Does the map function increase or decrease processing time? I hope to be processing a stream with the code. (I'm trying to determine the presence of 2 objects)

cs95 Over a year ago

@Hojo.Timberwolf The difference is "seriously, don't worry about it" small, but you can look at the alt just below it if you're not interested.

Collectives™ on Stack Overflow

Clone a NumPy array multiple times

1 Answer 1

5 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related