remap

Usage

Remaps an image to implement geometric transforms such as rotation, scaling, and distortion. It generates a transformed image by specifying locations of source image pixels in the destination image, and is widely used in fields such as image correction and stereo vision.

Interface

cv2.remap(src, map1, map2, interpolation=cv2.INTER_LINEAR, dst=None, borderMode=cv2.BORDER_CONSTANT, borderValue=None)

Parameters

Parameter	Description	Value Range	Input/Output
src	Input image.	Not null	Input
map1	2D array containing the X-coordinate of each pixel in the source image (or the first part of the pixel mapping matrix).	32-bit floating point number within the range of [0, width–1], where width is the width of the source image.	Input
map2	2D array containing the Y-coordinate of each pixel in the source image (or the second part of the pixel mapping matrix).	32-bit floating point number within the range of [0, height–1], where height is the height of the source image.	Input
interpolation	Interpolation mode. The default value is cv2.INTER_LINEAR. Other options include cv2.INTER_NEAREST and cv2.INTER_CUBIC.	cv2.INTER_LINEAR, cv2.INTER_NEAREST, cv2.INTER_CUBIC, etc.	Input
dst	(Optional) Output image. A new image is created by default.	Not null	Output
borderMode	Border mode. The default value is cv2.BORDER_CONSTANT which determines how to process pixels outside the source image.	Common modes: cv2.BORDER_CONSTANT, cv2.BORDER_REPLICATE, and cv2.BORDER_REFLECT	Input
borderValue	Border value, which is valid only when borderMode is set to BORDER_CONSTANT.	For a single-channel grayscale image, borderValue is usually an integer. For a multi-channel image, borderValue is a tuple containing multiple integer values.	Input

Return Value

Output image

Example

import numpy as np
import cv2

# Create a 5x5 image.
src = np.array([[0, 0, 0, 0, 0],
               [0, 1, 1, 1, 0],
               [0, 1, 0, 1, 0],
               [0, 1, 1, 1, 0],
               [0, 0, 0, 0, 0]], dtype=np.uint8)  

# Create a mapping matrix to define image remapping.
map_x = np.array([[0, 1, 2, 3, 4],
                 [0, 1, 2, 3, 4],
                 [0, 1, 2, 3, 4],
                 [0, 1, 2, 3, 4],
                 [0, 1, 2, 3, 4]], dtype=np.float32)  

map_y = np.array([[0, 0, 0, 0, 0],
                 [1, 1, 1, 1, 1],
                 [2, 2, 2, 2, 2],
                 [3, 3, 3, 3, 3],
                 [4, 4, 4, 4, 4]], dtype=np.float32)

dst = cv2.remap(src, map_x, map_y, cv2.INTER_LINEAR)

print(dst)

Output:

[[0 0 0 0 0]
 [0 1 1 1 0]
 [0 1 0 1 0]
 [0 1 1 1 0]
 [0 0 0 0 0]]

Parent topic: Interface Definition