Basic Cooking #1 -Open,Save, Resize, Display

OpenCV Open image

Use cv2.imread function to open an image file.

import cv2

img = cv2.imread("dog.jpg", cv2.IMREAD_COLOR)
height, width, channels = img.shape
print("JPG   H:%d W:%d, Channel:%d"%(height, width, channels))

#There's no channel value when open with grayscale
img = cv2.imread("dog.jpg", cv2.IMREAD_GRAYSCALE)
height, width = img.shape
print("JPG   H:%d W:%d "%(height, width))

img = cv2.imread("linux.png", cv2.IMREAD_COLOR)
height, width, channels = img.shape
print("PNG 1 H:%d W:%d, Channel:%d"%(height, width, channels))

img = cv2.imread("linux.png", cv2.IMREAD_UNCHANGED)
height, width, channels = img.shape
print("PNG 2 H:%d W:%d, Channel:%d"%(height, width, channels))
<open.py>

Flags

  • cv2.IMREAD_COLOR : BGR format image is opened. So the channel will be 3.
  • cv2.IMREAD_GRAYSCALE : only 1 grayscale channel is opened. So the channel informaion is not necessary and there's no channel value in the img.shape.
  • cv2.IMREAD_UNCHANGED : Retains the format stored in the image file. If there's an alpha channel in the image, BGRA format image is opened. So the channel will be 4. 


root@spypiggy-ubuntu:/usr/local/src/study/python/opencv# python3 open.py
JPG   H:174 W:290, Channel:3
JPG   H:174 W:290
PNG 1 H:395 W:386, Channel:3
PNG 2 H:395 W:386, Channel:4

Be Careful : If you open the image in grayscale mode or convert the color image to grayscale, the shape returns only height and width values because there is only one channel. Therefore, an error occurs when trying to read the channel value.


If you want to open online image, use urllib

import numpy as np
import urllib.request
import cv2

url = 'http://images.cocodataset.org/val2017/000000439715.jpg'
req = urllib.request.urlopen(url)
img = np.asarray(bytearray(req.read()), dtype=np.uint8)
img = cv2.imdecode(img, cv2.IMREAD_COLOR)
height, width, channels = img.shape
print("H:%d W:%d, Channel:%d"%(height, width, channels))


PIL Open image

Use PIL open function to open an image file.


from PIL import Image

img = Image.open("f:\\tmp\\outg.jpg")
w, h = img.size
print('W:%d H:%d'%(w, h))
<open_pil.py>

root@spypiggy-ubuntu:/usr/local/src/study/python/opencv# python3 open_pil.py
W:174 H:290

Be Careful : When using PIL, the size of the image is stored in Image.size. While OpenCV's Mat.shape is stored in the order of height, width and channel, the size of PIL is stored in the order of width and height. And channel information is not stored.

If you want to know the channel information, use Image.mode


from PIL import Image

img = Image.open("f:\\tmp\\outg.jpg")
w, h = img.size
print('W:%d H:%d'%(w, h))

print(img.mode)

The mode values ​​are defined as follows at https://pillow.readthedocs.io/en/5.1.x/handbook/concepts.html#modes.

The mode of an image defines the type and depth of a pixel in the image. The current release supports the following standard modes:
  • 1 (1-bit pixels, black and white, stored with one pixel per byte)
  • L (8-bit pixels, black and white)
  • P (8-bit pixels, mapped to any other mode using a color palette)
  • RGB (3x8-bit pixels, true color)
  • RGBA (4x8-bit pixels, true color with transparency mask)
  • CMYK (4x8-bit pixels, color separation)
  • YCbCr (3x8-bit pixels, color video format)
    • Note that this refers to the JPEG, and not the ITU-R BT.2020, standard
  • LAB (3x8-bit pixels, the L*a*b color space)
  • HSV (3x8-bit pixels, Hue, Saturation, Value color space)
  • I (32-bit signed integer pixels)
  • F (32-bit floating point pixels)


If you want to open online image, use requests

import numpy as np
import requests
from PIL import Image

url = 'http://images.cocodataset.org/val2017/000000439715.jpg'
img = Image.open(requests.get(url, stream=True).raw).convert('RGB')
w, h = img.size
print('W:%d H:%d'%(w, h))


OpenCV Image to numpy vice versa

OpenCV images (Mat) have a numpy array and can be used directly without conversion.


import cv2

img = cv2.imread("dog.jpg", cv2.IMREAD_COLOR)
print(type(img));




root@spypiggy-ubuntu:/usr/local/src/study/python/opencv# python3 shape.py
<class 'numpy.ndarray'>


Numpy arrays can be used as Mat objects in opencv without conversion.

 

import cv2  
import numpy as np 
img = np.zeros((512,512,3),np.uint8)
height, width, channels = img.shape

 

PIL Image to numpy vice versa

PIL uses image channels in the RGB order. However, OpenCV uses an image channel in the BGR order, so you need to  change the image format from RGB to BGR if the color image is opened.


from PIL import Image
im = Image.open(filename)
array = np.asarray(im, dtype="uint8")
array = cv2.cvtColor(array, cv2.COLOR_RGB2BGR) 

And you can convert numpy array to an PIL image using fromarray function(same as OpenCV).

from PIL import Image
im = Image.fromarray(np.uint8(array))
width, height = im.size

OpenCV Pixel Position

Common coordinates in mathematics use the (x, y) format. However, the example below shows that OpenCV has a shape value in the order of height, width, and channel. Therefore, OpenCV has (y, x, channel) coordinates.

import cv2  
import numpy as np 
img = np.zeros((512,100,3),np.uint8)
height, width, channels = img.shape

 

For example, if you have a png image of 2X2 pixels as shown in the following figure, you will have a shape of (22,24).

And the values of the four pixels are as follows. As the result shows (y, x, c).

import cv2
img = cv2.imread('c:\\lsh\\study\\opencv\\4pixel.png', cv2.IMREAD_UNCHANGED)
print(img[0][0])
print(img[1][0])
print(img[0][1])
print(img[1][1])

#### output ####
[255 255 255   0]
[255 255 255 255]
[  0   0   0 255]
[  0 216 255 255]

 

OpenCV Format conversion

Use cv2.cvtColor function to convert an image format.

import cv2


img = cv2.imread("linux.png", cv2.IMREAD_UNCHANGED )
height, width, channels = img.shape
print("RGBA H:%d W:%d, Channel:%d"%(height, width, channels));

#convert to gray
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
height, width = gray.shape
print("GRAY H:%d W:%d"%(height, width));

#remove alpha (you can also use cv2.COLOR_BGRA2RGB)
bgr = cv2.cvtColor(img, cv2.COLOR_BGRA2BGR)
height, width, channels = bgr.shape
print("BGR  H:%d W:%d, Channel:%d"%(height, width, channels));

#bgr -> rgb
rgb = cv2.cvtColor(bgr, cv2.COLOR_BGR2RGB)
height, width, channels = rgb.shape
print("RGB  H:%d W:%d, Channel:%d"%(height, width, channels));




root@spypiggy-ubuntu:/usr/local/src/study/python/opencv# python3 conv.py
RGBA H:395 W:386, Channel:4
GRAY H:395 W:386
BGR  H:395 W:386, Channel:3
RGB  H:395 W:386, Channel:3

The following values ​​are frequently used for conversion.
  • COLOR_BGR2BGRA = 0,
  • COLOR_RGB2RGBA = COLOR_BGR2BGRA,
  • COLOR_BGRA2BGR = 1,
  • COLOR_RGBA2RGB = COLOR_BGRA2BGR, 
  • COLOR_BGR2RGBA = 2,
  • COLOR_RGB2BGRA = COLOR_BGR2RGBA,
  • COLOR_RGBA2BGR = 3,
  • COLOR_BGRA2RGB = COLOR_RGBA2BGR,
  • COLOR_BGR2RGB = 4,
  • COLOR_RGB2BGR = COLOR_BGR2RGB,
  • COLOR_BGRA2RGBA = 5,
  • COLOR_RGBA2BGRA = COLOR_BGRA2RGBA,
  • COLOR_BGR2GRAY = 6,
  • COLOR_RGB2GRAY = 7,

PIL Format Conversion

Use Image.convert function to convert an image format. For usage, refer to the following document (https://pillow.readthedocs.io/en/stable/reference/Image.html).

Image.convert(mode=None, matrix=None, dither=None, palette=0, colors=256)[source]
Returns a converted copy of this image. For the “P” mode, this method translates pixels through the palette. If mode is omitted, a mode is chosen so that all information in the image and the palette can be represented without a palette.
The current version supports all possible conversions between “L”, “RGB” and “CMYK.” The matrix argument only supports “L” and “RGB”.
When translating a color image to greyscale (mode “L”), the library uses the ITU-R 601-2 luma transform:
L = R * 299/1000 + G * 587/1000 + B * 114/1000
The default method of converting a greyscale (“L”) or “RGB” image into a bilevel (mode “1”) image uses Floyd-Steinberg dither to approximate the original image luminosity levels. If dither is NONE, all values larger than 128 are set to 255 (white), all other values to 0 (black). To use other thresholds, use the point() method.
When converting from “RGBA” to “P” without a matrix argument, this passes the operation to quantize(), and dither and palette are ignored.
Parameters:
  • mode – The requested mode. See: Modes.
  • matrix – An optional conversion matrix. If given, this should be 4- or 12-tuple containing floating point values.
  • dither – Dithering method, used when converting from mode “RGB” to “P” or from “RGB” or “L” to “1”. Available methods are NONE or FLOYDSTEINBERG (default). Note that this is not used when matrix is supplied.
  • palette – Palette to use when converting from mode “RGB” to “P”. Available palettes are WEB or ADAPTIVE.
  • colors – Number of colors to use for the ADAPTIVE palette. Defaults to 256.
Return type:
Image
Returns:
An Image object.
The following example converts an RGB image (linearly calibrated according to ITU-R 709, using the D65 luminant) to the CIE XYZ color space:
rgb2xyz = (
    0.412453, 0.357580, 0.180423, 0,
    0.212671, 0.715160, 0.072169, 0,
    0.019334, 0.119193, 0.950227, 0)
out = im.convert("RGB", rgb2xyz)

OpenCV Resize

Use cv2.resize function to resize an image file. 


import cv2
img = cv2.imread('linux.png', cv2.IMREAD_UNCHANGED)
height, width, channels = img.shape
print("H:%d W:%d, Channel:%d"%(height, width, channels));

img2 = cv2.resize(img, (int(width / 2), int(height / 2)))
height, width, channels = img2.shape
print("H:%d W:%d, Channel:%d"%(height, width, channels));




root@spypiggy-ubuntu:/usr/local/src/study/python/opencv# python3 resize.py
H:395 W:386, Channel:4
H:197 W:193, Channel:4

PIL Resize

Use Image.resize function to resize an image file. 


img = Image.open('./linux.png')

width, height = img.size
print("H:%d W:%d"%(height, width));

img2 = img.resize((int(width / 2), int(height / 2)))
width, height = img2.size
print("H:%d W:%d"%(height, width));

OpenCV Display image

Use cv2.imshow function to display an image file.

Tips : If you use a remote ssh shell, use a ssh terminal like MobaXterm which supports X11 forwarding. Then you can view the display image windows.

First parameter is window name, and if there's multi windows, each window name must be different.
cv2.waitKey(time) function waits for keyboard input. If time value is 0, wait forever for keyboard input. If time value is not zero, it waits for time value(ms) and returns.


import cv2

img = cv2.imread("dog.jpg", cv2.IMREAD_COLOR)
height, width, channels = img.shape
print("JPG   H:%d W:%d, Channel:%d"%(height, width, channels));
cv2.imshow('color', img)

#There's no channel value when open with grayscale
img2 = cv2.imread("dog.jpg", cv2.IMREAD_GRAYSCALE)
height, width = img2.shape
print("JPG   H:%d W:%d "%(height, width));
cv2.imshow('gray', img2)

img3 = cv2.imread("linux.png", cv2.IMREAD_COLOR)
height, width, channels = img3.shape
print("PNG 1 H:%d W:%d, Channel:%d"%(height, width, channels));
cv2.imshow('png1', img3)

img4 = cv2.imread("linux.png", cv2.IMREAD_UNCHANGED)
height, width, channels = img4.shape
print("PNG 2 H:%d W:%d, Channel:%d"%(height, width, channels));
cv2.imshow('png2', img4)
cv2.waitKey(0)
cv2.destroyAllWindows()


Four image windows open like this.



PIL Display image

Use Image.show function to display an image file.


img = Image.open("f:\\tmp\\dog.jpg")
width, height = img.size
print("H:%d W:%d"%(height, width));
img.show()

img2 = Image.open("f:\\tmp\\dog.jpg")
img2 = img2.convert('LA')
img2.show()

Be Careful : PIL uses the operating system's default image viewer to display images.
Therefore, a new process is created. The show () function is returned only when the newly created image viewer process is terminated. Therefore, like OpenCV, multiple windows are not created at the same time. In the above example, after the first color dog.jpg picture window is opened, the corresponding image viewer program must be closed before the next grayscale dog image is displayed.

OpenCV Save image

Use cv2.imwrite function to save an image file.
This code will save 1/4 size of linux.png as small_linux.png.


import cv2
img = cv2.imread('linux.png', cv2.IMREAD_UNCHANGED)

height, width, channels = img.shape
print("H:%d W:%d, Channel:%d"%(height, width, channels));

img2 = cv2.resize(img, (int(width / 2), int(height / 2)))
height, width, channels = img2.shape
print("H:%d W:%d, Channel:%d"%(height, width, channels));
cv2.imwrite("./small_linux.png", img2)


PIL Save image

Use Image.save function to save an image file.
This code will save 1/4 size of linux.png as small_linux.png.


img = Image.open('./linux.png')

width, height = img.size
print("H:%d W:%d"%(height, width));

img2 = img.resize((int(width / 2), int(height / 2)))
width, height = img2.size
print("H:%d W:%d"%(height, width));
img2.save('./small_linux.png')



댓글

댓글 쓰기

이 블로그의 인기 게시물

Image Processing #7 - OpenCV Text

이미지
You can download the source codes here( https://github.com/raspberry-pi-maker/OpenCV ) Basic text output using putText putText function Draws a text string. Python: cv2. putText ( img, text, org, fontFace, fontScale, color [ , thickness [ , lineType [ , bottomLeftOrigin ] ] ] ) → None Python: cv. PutText ( img, text, org, font, color ) → None Parameters: img – Image. text – Text string to be drawn. org – Bottom-left corner of the text string in the image. font – CvFont structure initialized using InitFont() . fontFace – Font type. One of FONT_HERSHEY_SIMPLEX , FONT_HERSHEY_PLAIN , FONT_HERSHEY_DUPLEX , FONT_HERSHEY_COMPLEX , FONT_HERSHEY_TRIPLEX , FONT_HERSHEY_COMPLEX_SMALL , FONT_HERSHEY_SCRIPT_SIMPLEX , or FONT_HERSHEY_SCRIPT_COMPLEX , where each of the font ID’s can be combined with FONT_ITALIC to get the slanted letters. fontScale – Font scale factor that is multiplied by the font-specific base size. color – Text color. thickness – Th...
자세한 내용 보기

Playing YouTube videos using OpenCV

이미지
So far, we have dealt with many examples of playing video files such as mp4, avi, or frames captured using an input device such as a webcam. In this post, we will learn how to play YouTube videos using OpenCV. You can download the source codes here( https://github.com/raspberry-pi-maker/OpenCV ) prerequisite You only need to install the papy, youtube-dl packages. pip3 install papy pip3 install --upgrade youtube-dl Pafy Objects and Stream Objects Pafy objects relate to videos hosted on YouTube. They hold metadata such as title, viewcount, author and video ID. Stream objects relate to individual streams of a YouTube video. They hold stream-specific data such as resolution, bitrate and url. Each Pafy object contains multiple stream objects. If you use papy, you can get information about youtube videos as follows. import pafy url = 'https://www.youtube.com/watch?v=M-UKXCUI0rE' video = pafy . new(url) #The author of the video (str) print(video . author...
자세한 내용 보기

OpenCV Installation - Rasbian Buster, Jessie, DietPi Buster

이미지
In this post, we will install OpenCV on Raspberry Pi . Raspberry Pie doesn't have a GPU, so you don't need to set up acceleration with CUDA. Therefore, it is easiest to install a package that has already been created. But old version of Rasbian(Jessie) doesn't have python package. So you have to build the source code to install the OpenCV. And I'm going to install OpenCV 4.1.1 to DietPi. DietPi is a lightweight version of Debian. Prepare Rasbian Buster headless sd card image Prepare the image using Etcher. I used 2019-09-26-raspbian-buster-lite.zip image. Do not boot the Raspberry Pi immediately after creating the image, but create a size 0 ssh file on the / boot partition. The reason for this file is to make it a headless image. This allows remote ssh operations without having to connect a monitor to the Raspberry Pi. However, to get an IP address using dhcp, the network cable must be connected. <Create ssh file in the boot partition> Then run the...
자세한 내용 보기