英文摘要 | One of the prerequisites of integrated circuit analysis is to get the panoramic image of the chips’ internal structure. At present, it often needs to collect millions of clear microscopic images to analysis nano-scale IC chips. Therefore, it becomes an urgent task to establish a stable, accurate, and fast large scale IC microscopic images acquisition system based on the characteristics of IC chips. IC chips are of the three-dimensional structure of multilayer routing. To acquire every layer images, equipment are needed, including optical microscopes and scanning electron microscopes. However, the transparent oxide layer medium limits the usage of the auto-focusing algorithms in the optical microscope. For the SEM, when the depth of field can’t cover the fluctuation of the curved surface, manual monitoring is still needed to acquire accurate surface images. According to the problems proposed above, this study systematically analyzed the correlation among the features of the microscopic instruments, the surface characteristics and the movement accuracy of the motorized stage, then developed a microscopic images acquisition system using pre-acquisition images to monitor the acquisition process, and applied to acquire IC microscopic images. The dissertation solves three key technique issues: designing a pre-acquisition/ acquisition approach, that is, acquire the right pre-acquired image as a standard to monitor the acquired images, in order to reduce the accumulated error and improve the stability of the system; designing an image selection method based on image registration, to guarantee the accuracy of the acquisition process; designing a platform height forecast method based on surface fitting, to significantly improve the speed of acquisition process. Based on these methods and techniques, the dissertation developed an IC optical- microscopic image acquisition system based on surface fitting and image registration, as well as an IC electron-microscopic image acquisition system based on image registration, which can acquire image continuously, automatically and rapidly, and make the image acquisition process with high stability and high accuracy. The correct microscopic images of integrated circuit chips acquired by the system developed in this dissertation provide an important foundation for the integrated circuit analysis. One the other hand, the system also can be applied to the high-resolution three-dimensional structure imaging of other biologica... |
修改评论