<p>Fragmented reassembly is a fundamental and classic geometric processing problem with numerous applications in fields such as archaeology and forensics. However, most traditional methods, which rely on dynamic programming to solve the boundary matching problem, suffer from high time complexity and can be quite time-consuming. In this paper, we propose an effective and efficient method for the reassembly of 2D image fragments to address these challenges. First, we compute potential matches between pairs of image fragments using multi-resolution isometric sampling. Next, we employ a graph-based clustering method to reassemble the complete image. Finally, we utilize a graph-based optimization algorithm to refine the reassembly results. Extensive experiments on multiple datasets demonstrate that our algorithm outperforms the state-of-the-art, achieving superior performance with significantly reduced runtime.</p>