ImgLib2-based re-slicing viewer for terabyte-sized multi-view image sequences. See https://imagej.net/plugins/bdv for (somewhat out-dated) user documentation.

BigDataViewer is designed to be re-used as a visualization component in other software (besides being a stand-alone Fiji plugin). Examples include ABBA, BigStitcher, BigWarp, Labkit, MaMuT, Mastodon, MoBIE, Paintera.

vistools

The best entry point for using BDV in your own project is the BdvFunctions class. A short introduction is below, and you can find a lot of short examples here.

Usage

Random random = new Random();
Img<ARGBType> img = ArrayImgs.argbs(100, 100, 100);
img.forEach(t -> t.set(random.nextInt()));
Bdv bdv = BdvFunctions.show(img, "img");

creates a random 3D ARGB img and shows it in a BDV window.

All BdvFunctions methods will return some instance of Bdv which can be used to close the BDV window:

bdv.close()

or add more stuff to the same window:

BdvFunctions.show(img2, "img2", Bdv.options().addTo(bdv));

Via bdv.getBdvHandle() you can get access to the BDV ViewerPanel and SetupAssignments allowing more fine-grained manipulations of BDV state.

The concrete Bdv instance returned from BdvFunctions.show(img, "img") in the above example is BdvStackSource<ARGBType> and has additional methods, e.g., removeFromBdv() which removes img from the BDV window.

To display a 2D image, display a 3D image as a stack of 2D slices over time etc:

BdvFunctions.show(img2, "img2", Bdv.options().is2D());

The is2D() option is per Viewer window, not per stack. If it is set, the BDV navigation mouse and keybindings are set up for 2D, etc...

More fine-grained control can be achieved with axisOrder() option:

BdvFunctions.show(img2, "img2", Bdv.options().is2D().axisOrder(XYT))

AxisOrder specifies how the stack dimensions are interpreted. For BDV with 3D navigation, the follwoing are useful: XYZ, XYZC, XYZT, XYZCT, XYZTC. For BDV with 2D navigation, the following are useful: XY, XYC, XYT, XYCT, XYTC. You should interpret C and T losely. The effect is that T will be mapped to the time slider of the BDV, while C is mapped to BDV sources. (This also means that you should not have images with a large C dimension.)

There is/will be stuff to add annotation overlays. Currently BdvFunctions.showPoints() is available.

Here is an example screenshot where several 3D stacks, a set of boundary points, and 3D ellipsoids fitted to the boundary points were added to a BDV window. It also shows how the usual BDV dialogs can be used to adjust visibility / brightness / colors of tbe individual sources.