Characteristic Recognition of IC Chip’s Micro-Topography Defects Based on Image Projection Transformation and Energy Optimization Modeling
For accurate inspecting the micro-topography defects of IC chip, characteristic recognizing method of topography defects based on image projection transformation and energy optimization modeling is presented. First the topography conjugate images have been obtained by high-definition CCD cameras from different space positions, and then we extracting the relative positions of several selected IC defect characteristic-points, through establishing the relationship of image projecting transformation these point’s absolute three-dimensional coordinates can be calculated and determined, with energy optimization modeling the controlling curves are fitted, and then the defect topography in each enclosing spatial region which constituted by these controlling curves is structured. Mathematical characteristics are established according to the topography model, after training and adjusting BP network repeatedly the topography defects can be characteristic-recognized, and the recognizing results are gotten. In this experiment the performance comparison and analysis are implemented in several typical recognizing methods of image characteristic, which prove this new method’s accuracy and reliability and provide new idea for following IC chip’s precise detection.
Liangchi Zhang, Chunliang Zhang and Tielin Shi
Z. W. Liang et al., "Characteristic Recognition of IC Chip’s Micro-Topography Defects Based on Image Projection Transformation and Energy Optimization Modeling", Advanced Materials Research, Vols. 139-141, pp. 1990-1995, 2010