Microscopy based phenotypic analysis is a powerful technique for investigating biological mechanisms and molecular signalling networks. For effective analysis high throughput approaches are required to estimate the inherent variability of biological systems. Automated image analysis is essential for processing the large volumes of data involved in the assays. Image segmentation is a critical part of the image analysis but obtaining accurate and reliable automated segmentation remains a challenge. hi my research 1 have worked to address the problem of estimating the foreground of a fluorescent nucleus marker image by identifying relationships between clusters of pixels. This technique is proven to be more effective at identifying the image foregrounc than the standard approach of classifying pixels based upon intensity thresholds.