Blood cells are vital components of the circulatory system, playing critical roles in oxygen transport, immune defense, and clot formation. The morphology of white and red blood cells can reveal diseases and disorders, making accurate segmentation and classification essential for hematological diagnostics. Beside standard histologygocal cytolpgica appaoches, biophotonics is emering as new set of tools fur such diagnostics. These bio-photonic processes enable label-free imaging of unstained blood smears, leveraging intrinsic cellular properties such as morphology, refractive index, and texture. Unlike stained blood smear analysis, which relies on color differentiation, unstained analysis depends on intrinsic cell properties, presenting challenges such as low contrast, subtle feature variations, and imaging artifacts. This review evaluates rule-based, machine learning, and deep learning techniques for segmentation and classification of unstained samples of cells, highlighting their strengths, challenges, and potential to improve diagnostic accuracy, clinical applications, and innovations in biophotonics.