apetsiuk / microstructure-vision-based-porosity-analysis

Microstructure vision-based porosity analysis

Microstructure vision-based porosity analysis

Notebooks:

• porosity_analysis_aluminum.ipynb

Large scale image (Aluminum, 300 microns)

Input data:

• Grayscale threshold = 60
• Minimum pore area = 50 px2
• Maximum pore area = 50000 px2
• Critical eccentricity = 0.6

Output data:

• Image area = 2595840 px2
• Pore area = 36730 px2
• Pore fraction = 1.41%
• Total number of pores = 630
  • Number of circular pores = 45
  • Number of elliptical pores = 142
  • Number of ignored pores = 443
  • Total area of circular pores = 36730 px2
  • Total perimeter of circular pores = 2693 px
  • Median equivalent diameter of circular pores = 30 px
  • Mean eccentricity of circular pores = 0.48

Small scale image (Aluminum, 30 microns)

Input data:

• Grayscale thresholds = [0…44; 45…84; 85…255]

Output data:

• Source image area = 3145728 px2
• Total 'dark' area = 45145.0 px2 (1.44%)
• Total 'mid' area = 919080.0 px2 (29.22%)
• Total 'light' area = 1899140.0 px2 (60.37%)