Awesome

Saccade (Sophisticated Image + OpticalFlow Viewer)

A saccade (/səˈkɑːd/ sə-KAHD) is a quick and simultaneous movement of both eyes in the same direction.

This project is the missing scientific HDR image viewer for Linux. It is OpenGL-based and aims at providing an efficient image viewer with some sophisticated functions for comparing images. When dragging and zooming an image in one viewport, all other viewports synchronously zoom and shift in the same way, so you see exactly the same patch in all images.

Features

• OpenGL accelerated viewer using mip-mapping data structure
• synchronize multiple viewports when dragging and zooming within one viewport
• drag'n drop for open images
• crop regions from image
• keyboard short-cuts for all actions
• set marker on a specific pixel
• double-click on information in statusbar copies the values into the clipboard
• filewatcher (images will be reloaded when they change)
• interactive histogram widget which effects the image
• supported file formats: png jpg bmp ppm tif CR2 and many more
• helpful commands to arrange multiple windows
• multi-threaded loading and writing

Supports the following file formats:

• image: *.png *.jpg *.jpeg *.bmp *.ppm *.tif *.CR2 *.JPG *.JPEG, *.JPE
• optical-flow: *.flo

Synchronized view-ports

Hereby, each viewport consists of different image-layers. Some action might affect only the active layer across all viewports (global effects), e.g., zooming and dragging, e.g. setting the crop-region:

When manipulating the histogram, the changes can be applied to only the current active layer or all layers in the current viewport (holding the SHIFT key). Further, a marker can be set precisely on a specific pixel, and all other views share the marker at exactly the same location. It uses the FreeImage library to load images from different types:

Install from binary

The related binaries for the master branch are generated automatically and statically linked to Qt5.9.2 to reduce dependencies and file-size. The pre-compile binaries are available with all dependencies and are tested in a VM with plain Ubuntu 16.04 .

Install from Source

Requirements

The requirements (Ubuntu 16.04) are currently:

• Qt5.9
• CMake
• FreeImage library
• Google Log
• Google Flags

I plan to add CUDA for speeding up some functions.

Get dependencies and compile

The dockerfile lists all dependencies for building this project. Or follow the following guide:

Assuming you are on Ubuntu 16.04 getting all dependencies is done by

# install dependencies
sudo apt-get install libfreeimage3 libfreeimage-dev libgflags-dev libgoogle-glog-dev

The Ubuntu-packages contain an old version of Qt5. To get a recent Qt5 version (>=5.9) use the following commands (you might need to register for Qt).

cd /tmp
wget http://download.qt.io/official_releases/qt/5.9/5.9.1/qt-opensource-linux-x64-5.9.1.run
chmod +x qt-opensource-linux-x64-5.9.1.run
./qt-opensource-linux-x64-5.9.1.run

And finally just get and build this repository by

# clone this repo
git clone https://github.com/PatWie/saccade.git
cd saccade
mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j4

A copy and paste version to automate the GUI installer is described in the Dockerfile.

To add the saccade application to the unity-launcher, edit the saccade.desktop file by adjust the paths:

edit saccade.desktop
cp saccade.desktop $HOME/.local/share/applications/saccade.desktop

and you find the app icon in the Ubuntu search bar. When debugging the application, it might be helpful to start it with the flag --logtostderr 1 and build it with DCMAKE_BUILD_TYPE=Debug.

Keyboard Shortcuts

These are very likely to changed in the next versions.

General shortcuts

shortcuts for a single viewport (visible layer)

shortcuts for local effects (all layers in single viewport)

shortcuts for global effects (synchronized across all viewports)