isistools 0.4.0

Python-based review tools for ISIS3 coregistration workflows

---

format: html: theme: solar

isistools

Python-based review tools for ISIS3 coregistration workflows. Replaces Qmos and Qnet with modern, interactive visualization using HoloViews, Panel, and datashader.

Why?

ISIS's Qmos and Qnet have several pain points:

• Qmos requires level-1 unprojected images but displays footprints in map projection (confusing), and renders image content very slowly.
• Qnet requires level-2 map-projected images but displays them flipped in detector readout order (makes comparison difficult).
• Both use a color scheme where registered control points are nearly invisible while unregistered points are prominent — the opposite of what you want during review.

isistools fixes all of these by building on the Python geospatial stack.

Features

• Footprint map: Interactive map of image footprints using geopandas + hvplot, with hover info and click-to-select.
• Image viewer: Fast rasterized image display via rioxarray + datashader. Images always shown in correct (north-up) orientation.
• Control network overlay: Tie points with sensible colors (registered = bright green, unregistered = red, ignored = gray). Residual vectors. Point detail on click.
• Dual interface: Same code works in Jupyter notebooks and as standalone Panel apps in the browser.
• CLI: isistools mosaic, isistools tiepoints, isistools footprints commands that launch Panel apps.

Installation

pip install -e .

Usage

CLI

# Mosaic review (Qmos replacement)
isistools mosaic cubes.lis --cnet control.net

# Tiepoint review (Qnet replacement)
isistools tiepoints cubes.lis control.net

# Quick footprint map
isistools footprints cubes.lis

# Control network summary stats
isistools cnet-info control.net

Notebook

from isistools.apps.mosaic_review import MosaicReview

app = MosaicReview("cubes.lis", cnet_path="control.net")
app.panel()  # renders inline

from isistools.apps.tiepoint_review import TiepointReview

app = TiepointReview("cubes.lis", "control.net")
app.panel()

Low-level API

from isistools.io.footprints import load_footprints
from isistools.io.controlnet import load_cnet
from isistools.io.cubes import load_cube
from isistools.plotting.footprint_map import footprint_map
from isistools.plotting.image_viewer import image_plot, image_with_cnet
from isistools.plotting.cnet_overlay import cnet_to_geodataframe

# Load data
gdf = load_footprints("cubes.lis")
cnet = load_cnet("control.net")
da = load_cube("image.cub")

# Plot footprints
footprint_map(gdf)

# Plot image with control points
image_with_cnet(da, cnet, serial_number="MRO/HIRISE/...")

# Convert cnet to GeoDataFrame for map overlay
cnet_gdf = cnet_to_geodataframe(cnet)

Requirements

• Cubes must have footprintinit already run for footprint display
• Control networks in ISIS3 binary format (compatible with jigsaw)
• GDAL with ISIS3 driver support (for reading .cub files via rioxarray)

Dependencies

Core: geopandas, hvplot, holoviews, datashader, panel, rioxarray, plio, pvl, typer

License

MIT