preface

For the conversion of color space, we will introduce you through HSV actual combat. As for the conversion and application of other color Spaces, I will learn other knowledge separately in the follow-up. Starting with this article, we will study the geometry transformation of OpenCV.

For example, in the field of image processing, we will often encounter image scaling, rotation and other operations, especially for users who have used PS experience, for these operations must be handy, but in fact, the underlying code is to explain the knowledge behind. In this article we will introduce scaling.

The zoom

In OpenCV, it gives us a scaling function called cv2.resize(), which is defined as follows:

def resize(src, dsize, dst=None, fx=None, fy=None, interpolation=None) :
Copy the code

SRC: represents the original image that needs to be scaled

Dsize: The size of a scaled image

Fx: represents the horizontal scaling ratio

Fy: indicates the scale in the vertical direction

Interpolation: Interpolation method

The interpolation values are shown in the table:

In cv2.resize(), the size of the target image can be specified by either “dsize” or “fx,fy”.

When you specify dsize, whether you specify fx,fy or not, the dsize parameter determines the size of the target image, that is, dsize has the highest priority. The specific mathematical formula is as follows:

Width = (double) dszie. Width/SRC cols

Height = (double) dszie. Height/SRC. The rows

Note that the first argument to dsize is the number of columns, and the second argument is the number of rows, as opposed to shape.

When you specify the target image size with fx,fy, the math is as follows:

dsize=Size(round(fxsrc.cols),round(fysrc.rows))

As for the final parameter interpolation, it refers to assigning values to pixels that cannot be directly obtained by mapping during geometric processing of the image. For example, if you want to enlarge an image by 2 times in a certain size area, some images may be too small to fit some pixels missing, and some may be too large to fit some pixels missing. For these pixels, interpolation determines how to determine their values.

Dsize enables scaling

Now that we know all the parameters of the resize function in OpenCV, let’s implement a simple image scaling as follows:

import cv2

img = cv2.imread("4.jpg")
rows,cols=img.shape[:2]
size=(int(cols*2),int(rows*1))
result = cv2.resize(img, size)
cv2.imshow("img", img)
print(img.shape)
cv2.imshow("result", result)
print(result.shape)
cv2.waitKey()
cv2.destroyAllWindows()
Copy the code

Here, we first get the image’s length and width pixels using img.shape, and then enlarge the rows unchanged and the columns by 2 times. After running, we get the following image.

Fx,fy scale

We used dszie parameters to scale above, now we use FX,fy parameters to scale as shown below:

import cv2

img = cv2.imread("4.jpg")
result = cv2.resize(img, None,fx=2,fy=1)
cv2.imshow("img", img)
print(img.shape)
cv2.imshow("result", result)
print(result.shape)
cv2.waitKey()
cv2.destroyAllWindows()
Copy the code

This code implements the same effect as above, and the results are not shown. As you can see, fx and FY code will be much more concise.