Skip to main content
. 2023 Jan 31;13(3):511. doi: 10.3390/diagnostics13030511
Algorithm 1: Procedures for segmentation of thin blood smears of P. falciparum and P. vivax
Input: Ring, trophozoite, schizont, and gametocyte images
Output: Segmented images of P. falciparum and P. vivax
Procedures:
  1. Apply the MGCS contrast enhancement technique to the original malaria image.

  • 2.

    Extract the G colour component from the enhanced RGB image.

  • 3.

    Apply the selected segmentation technique that had been determined from the analysis of different segmentation techniques.

  • 4.

    Check the area of background region, Area[bg] that satisfies this condition,

  •             a.

    If the value of Area[bg] >= 450,000 pixels, apply the 3×3 pixels median filter.

  •             b.

    Else, go to step 1.

  • 5.

    Apply the seeded region growing area extraction algorithm.

  •             

    Note: Regions not in the ranges of 150–1000 and 3500–30,000 were labeled as non-malaria parasites and excluded from the segmented image during this process.

  • 6.

    Apply the 5 × 5 pixels minimum filter.

  • 7.

    Perform the 5 × 5 pixels of the erosion process.

  • 8.

    Check the area of the foreground region, Area[fg] that satisfies this condition,

  •             a.

    If the value of Area[fg] >= 9000 pixels, apply the seeded region growing area extraction algorithm

  •                
    Note: Regions not in the 3500–30,000 range were labeled as non-malaria parasites and excluded from the segmented image during this process.
  •             b.
    Colorize the segmented image based on the enhanced RGB image.
  • 9.

    End.