Link zu WSL Hauptseite Swiss Federal Institute for Forest, Snow and Landscape Research WSL


ROXAS is designed to allow efficient and flexible analysis of anatomical features in the xylem and other tissues including phloem, rays and also conifer needles. The following features are included (see overview in Fig. 1).

a) Automatic image processing

  • Image improvement for inhomogeneous and poor contrast (see Fig. 2 below)
  • Edge enhancement

b) Automatic cell and ring detection

  • Automatic recognition of annual rings and cells in large samples (linear and circular) with >100 annual rings and 1,000,000 cells
  • Cell detection and filtering based on color, size, shape & context
  • Automatic filters to remove particles within cells and protrusions from cell walls (see Fig. 3 below)

c) Efficient manual editing of cells and ring borders

  • All measurements visualized as editable vector overlays
  • Selecting areas of interest (AOI) / exclusion (AOE)
  • Toolbox for automatic cell filtering using complex criteria
  • Deleting, drawing, correcting, undoing tools of cell lumen outlines and ring borders
  • Quality controlling sorted list of largest and smallest cells

d) Data output

e) Batch processing

  • Batch processing functions for all relevant steps to avoid waiting time for the user (spatial calibration, project preparation and execution, etc.)

f) Customizing analysis

  • Large collection of pre-defined configurations for many samples (see ROXAS Configuration Library)
  • Adjusting and creating own configurations to specifics of samples
  • Customizing colors of vector overlays (cells, ring borders, AOE, AOI)

g) Additional features

  • Analysis of length (considering curvatures!), width and area of scanned conifer needles
  • Analysis of tissue such as rays (currently only in circular cross-sections)
  • User manual (115 p.)

ROXAS analysis overview
Fig. 1: Overview of the ROXAS analysis including automatic image processing, anatomical feature detection, manual editing, and automatic output calculation and saving to MS Excel and text files.
Contrast homogenization
Fig. 2: ROXAS automatically corrects inhomogeneous and poor contrast, which improves cell detection as illustrated in the bottom row. Pinus leucodermis. Scale bar: 100 μm. Adjusted from von Arx et al. 2016, FPS
Protrusion filter
Fig. 3: ROXAS includes filters (two right panels) to correct sample deficiencies such as ripped-off cell walls extending into the cell lumina (upper yellow arrow), which avoids erroneous measurements. The 'Protrusion filter' is selective and maintains natural concavities as pit pore widening (lower yellow arrow). Pinus sylvestris. Scale bar: 100 μm. Adjusted from von Arx et al. 2016, FPS