Visual examples
This page documents the functionality of DICOMautomaton through visual examples. This page also helps track how
functionality evolves over time, and what functionality currently is (or is not) available.
Main view with images, meshes, and plots active:
Contouring and image adjustment options:
Script writing, compiling, and execution:
Some operations to be added into scripts or run via command line:
Scripts are plain text files!
Drawing / image editing interface:
Real-time plotting, row- and column-profiles at the current mouse position:
Mesh viewing, wireframe:
Mesh viewing, solid:
The terminal viewer can provide rudimentary visualization in headless scenarios, including over ssh or on systems not without OpenGL capabilities.
Using ansi and unicode:
Without ansi support and using ASCII (great for ssh viewing in linux built-in terminal!):
Handles anisotropic pixel width/height, more-or-less displaying the correct aspect ratio
Same as above, but with Gaussian noise added using PerturbPixels operation:
The web interface exposes many of the operations available on the command line and interactive viewers, but on a remote server. Operations must be performed sequentially in a modal fashion.
DICOMautomaton CI binaries can be run on an Android phone (armv7) via Termux! It also works on similar emulated shells
and the Android shell too. Static continuous integration builds are powered by musl libc and generally run without
issues, though not all functionality is included.
A stress test of the SDL-based viewer with a mesh composed of 19 million triangles maintained a rate of 52 fps. The test mesh was extracted from an image array containing random noise using marching cubes. Exporting the mesh file in PLY format makes a 700 MB file. I couldn't find another program that would even open it. (EDIT: Meshlab can also open it, but only gets 1.5 fps!)
Here are animation examples showing the image auto-advance mechanism in the SDL viewer:
Inspecting a spatial deformable registration vector field that was converted to both a mesh representation and an image array:
Constructive-solid geometry using signed-distance functions. Basic shapes, Boolean operations, and some 'finishing' effects like edge rounding (erosion) are present.
Another example showing how the SDL viewer auto-advance feature can be used for animation (MR Angiography):