python_image

Simple image processing library for python

Setup

Install dependencies

# pip3 install numpy

# pip3 install matplotlib

Clone repository

git clone https://github.com/cb285/python_image/

Import

from python_image import * or import python_image

Numpy inheritance

python_image's Img type inherits from numpy's ndarray, which allows for compatability with a large number of complex functions provided by numpy. more information can be found here.

if you need to convert between an Img type and ndarray, use the view function:

Img to ndarray: img.view(np.ndarray)

ndarray to Img: img.view(Img)

Reading image files

imread(filename)

Reads a file into an image.

Parameters:

• filename: string, name of image file to read

Returns:

• Img, image created by reading file

Example: img = imread("strawberries.png")

Supported file types:

• png
• NetPBM Formats (pbm, pgm, ppm)

Writing image files

Img.imwrite(filename)

Writes an image to file.

Parameters:

• filename: string, name of file to create

Filename extension specifies filetype. Supports same file types as reading.

Example: img.imwrite("my_image.png")

Displaying images

Img.imshow(figure, block)

Displays an image in a window.

Parameters:

• figure: integer, figure number - change for displaying multiple images at one time (default: 1)
• block: bool, if True then function blocks until the window is closed (default: True)

Example: img.imshow(1, False)

Note: if blocking is disabled then all windows will close when the program finishes.

New image creation

Img(shape, dtype)

Creates a new image with specified shape and data type initially filled with zeros.

Parameters:

• shape: tuple, shape (dimensions) of created array (default: (1, 1))
• dtype: data-type (python or numpy), datatype of created image (default: uint8)

Returns:

• Img, new image filled with zeros

Pixel addressing

2D image

Img[y, x]

• y: vertical distance from top of image
• x: horizontal distance from left of image

3D image

Img[z, y, x]

• z: plane
  • 0 : Red/Hue
  • 1 : Green/Saturation
  • 2 : Blue/Intensity
• y: vertical distance from top of image
• x: horizontal distance from left of image

Getting pixel value

Example: value = img[4, 30]

Setting pixel value

Example: img[43, 12] = value

Padding

Pad types:

• PADTYPE_NONE: padding disabled, out-of-bounds addressing will cause an exception (default)
• PADTYPE_ZERO: out-of-bounds addressing will return zero
• PADTYPE_SAME: out-of-bounds addressing will return the nearest pixel's value

Setting pad type:

Img.padtype(padtype)

Parameters:

• padtype: padtype, string constant specifying desired pad type

Getting current pad type

Img.padtype()

Returns the current pad type of the image as a string constant.

Returns:

• string constant, current padtype

RGB & HSI conversion

Img.rgb2hsi()

Converts the image from RGB to HSI representation.

Returns:

• Img, hsi representation of RGB image

Example: hsi = rgb_img.rgb2hsi()

Img.hsi2rgb()

Converts the image from HSI to RGB representation.

Returns:

• Img, rgb representation of HSI image

2D convolution

Img.conv2(mask, padtype)

Performs in-place 2D convolution. Values will be in the range [0, 1]

Parameters:

• mask: Img, odd-dimensioned image to use as mask
• padtype: string constant, specifies padding to be used during convolution (defaults to image's current pad type)