Sample Data¶

VIPP 0.16.0a7 includes 34 deterministic microscopy-style samples, including a bimodal threshold gallery. They are available inside Image Source:

Image Source > Source = sample

New in 0.16.0a4

Eight aligned image-review inputs bring the sample inventory to 34: red/green 2D intensities, a 3D intensity and Boolean mask, time-series intensity and categorical labels, and an explicit RGB volume with a numeric synthetic 0–1 index. Open the paired review examples. The index is not a computed RACC result. These samples also have a separate napari sample menu entry, VIPP synthetic image-review samples.

New in 0.16.0a3

VIPP synthetic 2D template detection (TCYX) and VIPP synthetic 3D template detection (TCZYX) add known fixed-orientation patterns with seeded noise, a nearby pair, a clipped border copy and a deliberately absent site. Select T=1 and Repeated pattern (C=1). The 3D spacing is Z=1.5, Y=0.5, X=0.4 micrometres. See the detection examples.

VIPP synthetic 2D spot tracking (TCYX) adds moving spots, an empty frame and an ambiguous crossing. VIPP synthetic 3D label tracking (TZYX) adds anisotropic objects with changing frame-local labels, a missed observation and a later appearance. Their tracking examples show the complete detection/measurement and linking routes.

0.16.0a1 adds a sixteenth sample, VIPP synthetic measurement plots: 60 isolated ellipses in one calibrated image for the plotting example.

This is the normal route when building or adapting a workflow. The sample menu then lists every bundled sample without adding raw layers to napari first.

Napari's sample-data menu is still useful when you want to inspect raw sample layers directly:

File > Open Sample > VIPP synthetic microscopy samples

Bundled Samples¶

Sample	Axes	Purpose
`VIPP synthetic volume`	`ZYX`	Basic grayscale z-stack with sphere/tube-like structures.
`VIPP synthetic multichannel volume`	`CZYX`	Three-channel fluorescence-like volume with blue/DAPI-like nuclei, green/FITC-like neurites, and red/TRITC-like puncta for channel splitting and segmentation.
`VIPP synthetic time-lapse multichannel`	`TCZYX`	Preferred starter sample with time, channel, z, y, and x axes.
`VIPP synthetic measurement summary`	`TYX`	Known object counts/areas per timepoint for summary validation.
`VIPP synthetic object morphology`	`YX`	Circle, ellipse, rectangle, and concave 2D objects for derived morphology.
`VIPP synthetic measurement plots`	`YX`	New in 0.16.0a1: 60 separated ellipses at 0.5 micrometer/pixel, with varied size/elongation and a deliberately generated size–brightness association.
`VIPP synthetic 3D mesh morphology`	`ZYX`	Anisotropic 3D objects for mesh volume, surface, convex hull, and sphericity checks.
`VIPP synthetic skeleton network`	`ZYX`	Sparse 3D skeleton-style network with known endpoints, junctions, spur, fragment, and isolated voxel.
`VIPP synthetic advanced skeleton network`	`TZYX`	Multi-timepoint skeleton stress test with loops, fragments, graph tables, and anisotropic scale.
`VIPP synthetic colocalization`	`CZYX`	Two-channel partial-overlap structures for colocalization, RACC-like output, and object association.
`VIPP synthetic deconvolution image`	`YX`	2D blurred/noisy image for PSF-aware restoration review.
`VIPP synthetic measured PSF`	`YX`	2D measured-PSF-like kernel for PSF preparation.
`VIPP synthetic 3D deconvolution volume`	`ZYX`	3D blurred/noisy volume for volumetric restoration review.
`VIPP synthetic 3D measured PSF`	`ZYX`	3D measured-PSF-like kernel for volumetric deconvolution.
`VIPP synthetic GPU segmentation cleanup`	`CZYX`	Four 3D objects, one removable 19-voxel speck, and one enclosed 31-voxel cavity for the portable GPU segmentation and Boolean mask-cleanup example.
`VIPP synthetic threshold gallery`	`ZYX`	Compact noisy phantom with separated background and foreground populations for global/local threshold inspection, including Minimum Threshold.
`VIPP synthetic 2D template detection`	`TCYX`	New in 0.16.0a3: known repeated patterns, a nearby pair, incomplete border copy and an absent site. Select T=1 and C=1.
`VIPP synthetic 3D template detection`	`TCZYX`	New in 0.16.0a3: volumetric patterns with anisotropic sampling and known calibrated centers. Select T=1 and C=1.
`VIPP synthetic 2D spot tracking`	`TCYX`	New in 0.16.0a3: moving spots, an empty frame and a crossing that is ambiguous for position-only linking.
`VIPP synthetic 3D label tracking`	`TZYX`	New in 0.16.0a3: anisotropic objects with changing frame-local labels, a missed observation and a later appearance.
`VIPP review 2D red intensity` / `VIPP review 2D green intensity`	`YX`	New in 0.16.0a4: aligned scalar channels for linked and additive-overlay review.
`VIPP review 3D intensity` / `VIPP review 3D mask`	`ZYX`	New in 0.16.0a4: anisotropic intensity and Boolean mask containing two authored physical spheres.
`VIPP review time-series intensity` / `VIPP review time-series labels`	`TZYX`	New in 0.16.0a4: four shared timepoints with categorical IDs 7 and 42.
`VIPP review 3D RGB composite` / `VIPP review 3D synthetic index`	`ZYXC` / `ZYX`	New in 0.16.0a4: explicit RGB with a constructed 0–1 numeric display phantom, not a RACC result.

Registration samples — new in 0.16.0a2¶

0.16.0a2 adds six registration samples: reference and moving images for a 2D subpixel pair, reference and moving volumes for a 3D rigid pair, a two-channel XYZ drift time series, and its companion label series. They are available through Image Source and the separate napari sample-data entry VIPP synthetic registration samples.

Use the registration examples for their connected workflows and known synthetic motion. Together with the 16 original microscopy samples and four detection/tracking samples above, these form the earlier 26-sample catalogue in 0.16.0a3. The eight image-review inputs added in 0.16.0a4 bring the current total to 34.

Documentation Samples Versus Core Samples¶

Core samples should stay small, deterministic, testable, and broadly useful.

Documentation-only samples can be more tutorial-specific. A documentation sample can become a core sample when it supports a stable workflow and has test coverage.

Synthetic truth is valuable for demonstrations and regression checks, but it does not establish performance on a biological assay. Validate a transferred workflow on representative real data and independent references.