Automatic Malaria Diagnosis in Digital Blood Smear Images
McMenamin, Katie
An automatic, image processing based diagnostic tool for Malaria could help fight the disease by making diagnosis cheaper, faster and more reliable. The most difficult step in the image processing pipeline is blood cell segmentation, with its accuracy limiting the efficacy of the entire algorithm. I present both pre-processing and post-segmentation methods for improving segmentation performance by addressing two main challenges: missing cells and under-segmenting clumped cells.
↧