jnhansen / vimg

A CLI image viewer

vimg: A CLI image viewer.

Author: Johannes Hansen

Why?

If you are like me, you spend a lot of time in command line environments. One of the few things that are hard to deal with in such an environment is images.

This little tool is meant to allow for quick viewing of image files in the command line.

The image is rendered using a combination of background color, foreground color and unicode character for each character cell to optimally represent the original pixels. The challenge is the limited color palette of 256 colors.

In the simplest case, each character cell represents one pixel. However, the resolution can be increased by printing unicode characters that better capture the structure of the image. Alternatively, the color accuracy can be improved by mixing two available colors in foreground and background, thus losing the gained resolution.

The default mode attempts to optimize the rendering by optimizing resolution in areas of high detail, and optimizing color accuracy in areas of low detail.

Installation

$ pip install vimg

Requirements

A terminal that supports 256 colors.

The script is based on curses and opencv for Python.

Usage

$ vimg path/to/image

GUI modes

Once in the GUI, you can change between different viewing modes:

Key shortcut	Mode	Description
c	color	(default) display optimal representation of image
f	fast	display image in fast mode (reduced resolution)
a	ascii	display a black-and-white representation of the image
e	edge	(experimental) edge detection based rendering
+/-	--	zoom in/out (by 30%)
h j k l or arrow keys	--	move view (by 10%)
0	--	reset zoom
r	--	refresh the screen
q	--	quit

Notes

The results will be better if you use a font that correctly displays unicode block element characters with the full line height, such as DejaVu Sans.

Limitations

The script currently only supports image files that are natively supported by OpenCV (.jpg, .png, .bmp).

To Do

Future plans include:

• Support for more image file types, e.g. .gif
• Improvement of the edge detection mode
• Make opencv dependency optional
• Improve color gradients at contrast-rich edges