Medial Axis Analysis of Three Dimensional, BiPhase Images

W. B. Lindquist
Department of Applied Mathematics and Statistics
State University of New York at Stony Brook, Stony Brook, NY 11974-3600
(516) 632-8361 (phone), (516) 632-8490 (fax)

September 30, 1999

Abstract:

This manual is intended to provide a guide for users of the software package 3DMA used to provide quantitative geometric analysis of 2D and 3D images of stochastic biphase materials.

• Contents
• Introduction
• Unbundling the Software Distribution
• Compiling 3DMA
• Testing the Compilation
  • Special Notes for Linux (Intel), Mac, and DEC Alpha Machines
• Execution
• 3DMA Organization
  • The Main Menu
  • Input Data Formats
    • Data File Name Convention
    • Supported User Formats
    • Segmented Data
    • Erosion Distance (`Burn') Data
    • Medial Axis Data
• Output Graphics Data
• Main Menu: Raw/Diffused Data
  • Segmentation
    • Global Thresholding
    • Diffusion
    • Kriging-Based Segmentation
    • Mardia-Hainsworth Segmentation Algorithm
    • Color Values
• Main Menu: Segmented Data
• Main Menu: Erosion Distance (`Burn') Data
• Main Menu: Medial Axis Data
• Sample Input Files
  • General Overview
  • Main Menu 1: Operations performed on Tomographic Data
    • case1.1: Generate Images of Input Data
    • case1.1_ras: Generate Rasterfile Images
    • case1.1a_ras: Generate Rasterfile Images with Fiducial Polygon
    • case1.1g_ras: Generate Rasterfile Images without Fiducial Polygon
    • case1.1_expl: Generate Iris Explorer 3D Image File
    • case1.2: Generate Attenuation Coefficient Histograms
    • case1.2/case1.2a.in: Generate Attenuation Coefficient Histograms with Fiducial Polygon
    • case1.2/case1.2g.in: Generate Attenuation Coefficient Histograms without Fiducial Polygon
    • case1.2/case1.2l.in: Generate Attenuation Coefficient Histograms without Fiducial Polygon
    • case1.3: Extract a Rectangular Subvolume of a Data Set
    • case1.4: Generate Circular Fiducial Polygon
    • case1.5dfn.slc: Segment Data using Anisotropic Diffusion Method on each Slice Separately
    • case1.5dfn.vol: Segment Data using Anisotropic Diffusion Method on 3D Data Set
    • case1.5krig: Segment Data Using Indicator Kriging
    • case1.5krig/case1.5ent.in: Indicator Kriging on Data Slices with Threshold Window Chosen by Entropy Function
    • case1.5krig/case1.5ent3d.in: Indicator Kriging on 3D Data Set with Threshold Window Chosen by Entropy Function
    • case1.5krig/case1.5mix.in: Indicator Kriging on Data Slices with Threshold Window Chosen by Binormal Mixture Fit
    • case1.5krig/case1.5mix3d.in: Indicator Kriging on 3D Data Set with Threshold Window Chosen by Binormal Mixture Fit
    • case1.5mh: Segment Data Using Mardia-Hainsworth Algorithm
    • case1.5mh/case1.5mh.in: Mardia-Hainsworth Segmentation on Data Slices
    • case1.5mh/case1.5mh.in: Mardia-Hainsworth Segmentation on 3D Data Set
    • case1.5smpl: Segment Data Using Simple Threshold
    • case1.6: Investigate Porosity Dependence of Threshold Choice
    • case1.6/case1.6a: Porosity vs. Threshold: File-by-File Scaling
    • case1.6/case1.6g: Porosity vs. Threshold: Global Scaling
    • case1.6/case1.6gu: Porosity vs. Threshold: Global Scaling with User Selected Range
    • case1.7: Investigate Free Parameter in Anisotropic Diffusion Method
    • case1.8: Investigate Application of Threshold Value on Diffused Images.
    • case1.8/case1.8s: Single threshold applied to slice diffused with multiple K values
    • case1.8/case1.8md: Multiple thresholds applied to single K slice
    • case1.8/case1.8mr: Multiple thresholds applied to a single slice of the original data set
    • case1.9: Plot Gradient of Image Intensity
    • case1.10: Compress Slices
  • Main Menu 2: Analyses Performed on Segmented Data Sets
    • case2.1: Generate Burn Data and Print Raster Images with Fiducial Polygon
    • case2.2: Generate Burn Data and Compute Medial Axis using Lee-Kashyap-Chu Algorithm
    • case2.2ms: Generate Burn Data and Compute Medial Surface using Lee-Kashyap-Chu Algorithm
    • case2.3: Compute Disconnected Volume Distributions
    • case2.3/case2.3.in: Disconnected Volume Distributions Using Erosion Algorithm
    • case2.3/case2.3.ntrim: Disconnected Volume Distribution Using Tree Search: No Edge Trimming
    • case2.3/case2.3.trim: Disconnected Volume Distributions Using Tree Search: Edge Trimming
    • case2.4: Compute Moment of Inertia Analysis on Disconnected Components
    • case2.5: Generate Images of the Segmented Data
    • case2.5_ras/case2.5_ras.in: SUN Rasterfile Format
    • case2.5_ras/case2.5a_ras.in: SUN Rasterfile Format: Convert Isolated Voxels
    • case2.5_BR: Bio-Rad MRC-600 Format
    • case2.5_dx: IBM Explorer Format
    • case2.5_expl: IRIS Explorer Format
    • case2.5_gl: Open GL Format
    • case2.5_iv: Inventor Format
    • case2.6: Convert Segmented Data Files to/from ASCII
    • case2.6/case2.6.ab.in: Convert ASCII Files to Binary
    • case2.6/case2.6.ba.in: Convert Binary Files to ASCII
    • case2.7: Compute Two Point Correlation Function
    • case2.7/case2.7.ci2.in: Grain - Classical 2D
    • case2.7/case2.7.grain.in: Grain - Classical x and y
    • case2.7/case2.7.m1d.in: Grain - Moving Window x
    • case2.7/case2.7.m2d.in: Grain - Moving Window 2D
    • case2.7/case2.7.m3df.in: Grain - Moving Window 3D Fast Algorithm
    • case2.7/case2.7.m3ds.in: Grain - Moving Window 3D Slow Algorithm
    • case2.7/case2.7.void.in: Void Phase - Classical x and y
    • case2.8: Comparison of 2-Pt Correlation Functions for Tomographic and Segmented Data Sets
    • case2.9: Compress Slices
    • case2.9/case2.9max.in: Compress Using Maximum Value
    • case2.9/case2.9maj.in: Compress Using Majority Value
    • case2.10: Shell Densities
    • case2.10/case2.10_cyl.in: Cylindrical Shells
    • case2.10/case2.10_sph.in: Spherical Shells
  • Main Menu 3: Analyses on Burn Data Sets
    • case3.1: Resize Burn Data Files
    • case3.2: Histogram and Attempt Fits to Burn Data Histogram
    • case3.2/case3.2.efit: Exponential Fit to Burn Data Histogram
    • case3.2/case3.2.nfit: Normal Fit to Burn Data Histogram
    • case3.3: Compute Medial Surface via Burn Algorithm
    • case3.4: Compute Medial Axis/Surface via LKC Algorithm
    • case3.4/case3.4.in: Medial Axis, top/bottom image boundaries represent truncated data set, original LKC sweep algorithm
    • case3.4/case3.4a12.in: Medial Axis, all boundaries capture full image, Stopy Brook sweep algorithm
    • case3.4/case3.4s.in: Compute Medial Surface via LKC Algorithm
    • case3.5: Convert Burn Data Files to/from ASCII
    • case3.5/case3.5.in: Convert Binary Files to ASCII
  • Main Menu 4: Analyses Available on Medial Axis Data Set
    • case4.1: Generate Images of the Medial Axis/Surface
    • case4.1_ras: SUN Rasterfile Format
    • case4.1_ras/case4.1a_ras.in: SUN Rasterfile Format
    • case4.1_ras/case4.1b_ras.in: SUN Rasterfile Format: Trim Axis, Add Grains
    • case4.1_ras/case4.1c_ras.in: SUN Rasterfile Format: Trim Axis
    • case4.1_BR: Bio-Rad MRC-600 Format
    • case4.1_dx: IBM Explorer Format
    • case4.1_dx/case4.1_dx.in: IBM Explorer Format - Image Entire Volume
    • case4.1_dx/case4.1a_dx.in: IBM Explorer Format - Image SubVolume
    • case4.1_expl: Iris Explorer Format
    • case4.1_expl/case4.1_expl.in: Iris Explorer Format - Image Medial Axis
    • case4.1_expl/case4.1s_expl.in: Iris Explorer Format - Image Medial Surface
    • case4.1_gl: Open GL Format
    • case4.1_gl/case4.1_gl.in: Open GL Format - Image Entire Volume
    • case4.1_gl/case4.1a_gl.in: Open GL Format - Image SubVolume
    • case4.1_iv: Inventor Format
    • case4.1_iv/case4.1_iv_a.in: Inventor Format - Image Volume
    • case4.1_iv/case4.1_iv_b.in: Inventor Format - Image Volume - offset origin
    • case4.2: Plot Medial Axis on Medial Surface - Iris Explorer Format
    • case4.3: Convert Medial Surface Files to Segmented File Type
    • case4.4: Histogram burn data for medial axis voxels only
    • case4.4/case4.4.ntrim.nper: No Axis Trimming
    • case4.4/case4.4.ntrim.per: No Axis Trimming, y-axis as percentage
    • case4.4/case4.4.trim.nper: Axis Trimming
    • case4.4/case4.4.trim.per: Axis Trimming, y-axis as percentage
    • case4.5: Embed Search Tree, Generate Histogram
    • case4.5/case4.5.ntrim: No Axis Trimming
    • case4.5/case4.5.trim: Axis Trimming
    • case4.6: Compute shortest path tortuosity distribution
    • case4.6/case4.6_34.in: Between faces 3 and 4 (x direction)
    • case4.6/case4.6_36.in: Between faces 3 and 6 (adjoining x-y faces)
    • case4.6/case4.6_56.in: Between faces 5 and 6 (y direction)
    • case4.7: Min Cut/Max Flow - Under Development
    • case4.8: Cluster Analysis - Test file not available
    • case4.9: Convert Medial Axis Data Files to/from ASCII
    • case4.9/case4.9.in: Convert Binary Files to ASCII
    • case4.10: Convert between Medial Axis Data Formats
    • case4.10cp/case4.10cp.in: voxel-list to cluster-path
    • case4.10g/case4.10g.in: voxel-list to graph-node
    • case4.11: Modify Medial Axis
    • case4.11v/case4.11v.in: output voxel-list format
    • case4.11cp/case4.11cp.in: output cluster-path format
    • case4.11g/case4.11g.in: output graph-node format
    • case4.12: Histogram Medial Axis Statistics
    • case4.12v/case4.12v.in: Input is voxel-list format
    • case4.12cp/case4.12cp.in: Input is cluster-path format
    • case4.13: Compute Throat Size Distribution
    • case4.13/case4.13_3sph.in: Use 3 sphere test data set
    • case4.13/case4.13_nl.in: Use 5 sphere test data set
    • case4.14: Identify Fibers
    • case4.15: Convert Cluster-Path Format to Graph-Node
    • case4.16: Construct Neuron - Under development
    • case4.17: Reconstruct Burn Layers - Under development
    • case4.18: Isolate Neuron Soma/Dendrite Structures - Under development
• Acknowledgements
• About this document ...