DART — Dendro Analysis & Reporting Tool

Beta

A specialized desktop workbench for the manual annotation and quantitative analysis of high-resolution tree-ring imagery.

DART replaces caliper-based manual measuring with a precise digital environment for “on-screen” dendrochronology, handling multi-gigabyte archival scans that typically overwhelm standard image viewers. Sub-pixel calibration, vector annotation, and standard RWL/CSV exports for downstream crossdating and chronology toolchains.

Stack Electron Node.js Sharp / libvips OpenSeadragon Fabric.js

DART: ChipLab

Beta

A desktop application that detects, measures, and species-classifies woodchips from flatbed scanner images — fully offline.

A companion to DART for wood-products research and mill QA. ChipLab converts a single flatbed scan into per-chip geometry (length, width, area, aspect ratio) and per-chip species predictions with confidence scores, using Mask R-CNN segmentation with SAM2 human-in-the-loop correction. Runs entirely offline; no data leaves the machine.

Stack Electron Python Mask R-CNN (MMDetection) SAM2 EfficientNet / ResNet-50

FOCUS — Flatbed Optical Capture & Utility Suite

In Development

A Windows application for driving an Epson Expression 13000XL flatbed scanner (with Transparency Unit) for bulk archival and scientific digitization of photographs, photo albums, dendrochronology specimens, and film.

Built for archival color fidelity and scientific measurement validity: IT8-target color calibration with cross-session drift tracking for both reflective and transparency light paths, embedded/sidecar metadata capture (manual, barcode, or batch CSV), and bulk scanning with hierarchical page/item auto-split and an operator review step before anything is finalized.

Stack .NET 10 / C# WPF NAPS2.Sdk Argyll CMS

IFML Archive Manager

Alpha

A custom Python-based toolkit for running the digitization workflow.

Automates the ingestion of high-resolution TIFF scans, preserves documented color and metadata attributes, and generates web-accessible Access Packages for deposit on Zenodo.

Stack Python 3 OpenCV Pillow (PIL) Zenodo API

Software Validity & Reproducibility

Guidance for Use

  • Validity: IFML software is provided as-is for research support. Outputs may contain errors and should be independently validated prior to publication or decision-critical use.
  • Reproducibility: Outputs may vary across operating systems, dependency versions, calibration choices, and model weights. Document software versions and dataset identifiers in any derived work.
  • System Requirements: Standard desktop environment (Windows 10+ / macOS 12+), 8 GB+ RAM, SSD storage recommended.