Home

MulimgViewer

1. Introduction

MulimgViewer is an image viewer, which is capable of displaying multiple images in one window. It is very useful for image comparison, selection and stitching.

Features:

• Loading images from multiple folders and displaying them side by side
• Simultaneous magnification of multiple selected areas
• Image rotation
• Remote directory support
• Bulk image resizing

Download link: Windows 10 binary package (amd64)

Quick start

Your star is my biggest support in developing and improving this project!

2. Python source code usage (windows\linux\ios)

After v3.9.1 only Windows 10 binary package (amd64) will be provided. But you can still use the source code to run the tool on other platforms. It is recommended to use Python 3.6 or newer.

pip installation: (If there is an error in the installation process, you can use conda.)

pip install wxpython pillow pytest-shutil numpy requests piexif

run:

python3 MulimgViewer.py

3. Windows 10

Execute the provided binary file directly. Or download the source code and run MulimgViewer.py.

After v3.9.3, Windows 10 provides installation version _Setup.exe and portable version _Portable.exe (installation version _Setup.exe starts faster).

4. Application Scenarios

Example 1: Parallel image browsing

How long does it take to browse a dataset - CelebA of 202,599 images? Display 1000 images at once using MulimgViewer — you will be all set in 200 clicks! See details.

Example 2: Parallel image selection and saving

Previously, you had to open each image separately, compare them one by one, then find the specific images in a folder and copy them to another place.

Now you can use MulimgViewer to load and display all the images from multiple directories at once, then export the image pairs with one click!

In Select output mode images will be copied by default, and moved if MoveFile option is selected. Use Select output mode with Parallel Auto or Parallel Manual input modes and disabled Parallel+Sequential option.

Example 3: Parallel zoom

MulimgViewer supports automatic image stitching, saving and magnification in parallel!

You can select any part of the image to magnify that area. Click Zoom 🔍, then click the left mouse button and draw a rectangle. Also you can duplicate the selection (click the right mouse button) and view all magnified areas simultaneously. See details.

Example 4: Dataset browsing, paired images

MulimgViewer makes it easy to view and compare paired images.

Example 5: One-click generation of paper comparison chart

Support to display the title and adjust the box position. When the box position is middle bottom, it is recommended to 🔍️Scale=-1,-1; if you choose other positions mode, adjust 🔍️Scale by yourself, for example: 🔍️Scale=1.5,1.5. See details

Show the half of an image. Check OneImg. When NumPerimg=2, combine two images into one image. When Gap(x,y)=*,*,0,*,*, the gap is eliminated.

Example 6: Browse images in a remote directory

After mounting a remote directory you can open that directory in MulimgViewer to view the images.

1. Ubuntu: In the example below we connected to server stfp://10.8.0.4 using Ubuntu's file manager nautilus.

2. win10: After installing WinFsp and SSHFS-Win, fill in the remote server ip in the file explorer, for example: \\sshfs\user@ip!port

Example 7: Click to rotate

Click Rotate ↷, then click the left mouse button to rotate an image.

Example 8: Batch image resizing

It is possible to batch resize images using the auto-save function. Step-by-step instructions:

1. Input mode: Sequential. Choose the input directory;
2. Choose the output directory;
3. Check AutoSaveAll;
4. Set TruthResolution to a fixed size, for example 256,256;
5. Click Save 💾️.