Sequence Metadata#

The tool //tools:ncore_sequence_meta extracts comprehensive metadata from an NCore sequences and outputs these as JSON file that references the component stores. The output includes the sequence ID, timestamp range, component store paths with MD5 checksums, component versions, and generic metadata fields.

This is useful when a data converter did not produce a metadata file, or when additional components have been added to an existing sequence and the metadata file needs to be regenerated to reflect all component stores.

Usage#

Basic invocation:

bazel run //tools:ncore_sequence_meta \
    -- \
    --output-dir=<OUTPUT_FOLDER> \
    v4 \
    --component-group=<SEQUENCE_META.json>

With multiple component groups (e.g., after extending a sequence with new components):

bazel run //tools:ncore_sequence_meta \
    -- \
    --output-dir=<OUTPUT_FOLDER> \
    v4 \
    --component-group=<COMPONENT_GROUP0> \
    --component-group=<COMPONENT_GROUP1>

Options#

Global options:

Option	Default	Description
`--output-dir`	(required)	Directory for the output JSON file
`--output-file`	`<sequence_id>.json`	Custom output filename
`--open-consolidated / --no-open-consolidated`	enabled	Pre-load consolidated zarr metadata for faster access
`--debug`	off	Enable debug-level logging

V4 sub-command options:

Option	Default	Description
`--component-group`	(required)	Component group or sequence meta path (repeatable)
`--poses-component-group`	`default`	Component group name for poses
`--intrinsics-component-group`	`default`	Component group name for intrinsics
`--masks-component-group`	`default`	Component group name for masks (`none` to disable)
`--cuboids-component-group`	`default`	Component group name for cuboids (`none` to disable)