Synthetic patterns imitating defect formation (formation of bridges or changes of thickness lines) during lithorgraphy

This dataset provides a collection of synthetic binary mask images  with  specifically designed for the evaluation of topology-aware metrics in semiconductor lithography image analysis. It includes images simulating two key defect types: (1) structural disconnections such as gaps and bridges, and (2) morphological deformations such as line width narrowing or thickening. Each image is generated from a randomized grid-based topology of manhatten patterns with controlled defects and is intended to support research in defect detection, graph-based image metrics, and topological analysis.