This work addresses the computational bottleneck in single-cell fluorescence microscopy image segmentation, which is essential for studying DNA damage tolerance pathways and cellular responses in health and disease. Existing tools such as MicrobeTracker are computationally intensive, require substantial manual input, and perform poorly on complex or densely packed cells.

To address these limitations, we propose a hierarchical image segmentation framework based on an iterative thresholding approach derived from the Chan–Vese model. The method adopts a coarse-to-fine multiscale strategy for fluorescence microscopy images. Specifically, an initial coarse segmentation is performed to separate the background, followed by a refined stage that accurately delineates cell regions and boundaries. Our approach reduces computational complexity and human intervention while maintaining high segmentation accuracy. Furthermore, we introduce a localized scheme via spatial decomposition to improve performance under non-uniform illumination.