| Improving scientific data visualization is important in study, data presentation, and publication. In Transmission Electron Microscope (TEM) applications, higher quality images are always expected; any noise appearing on a TEM image should be cleaned, but ultra-structural information needs to be well preserved during the image restoration procedure.; This thesis focuses on dealing with a chatter problem: alternating bright and dark bands overlapping within ultra-structure in a TEM image. The main goal pursued is not only to remove chatter bands from a chatter image but also to present evaluation results to show that the de-chatter images are acceptable because of the concern in scientific research for information loss.; Digital image processing techniques are employed to solve the chatter problem that is described in detail. Capitalizing on the periodic property of chatter, the blob-filter and Fourier transforms are used in the de-chatter operation. A suitable quantitative image quality metric, a universal image quality index (UIQI) is employed to assess the quality of a processed image. The analysis of filtering effects and de-chatter image quality presents the requirement of a balance between de-chatter and information loss. The results, which are acquired from image quality measurement by comparison of both chatter and de-chatter images following visual inspection and quantitative counting, demonstrate that a resulting optimum image can be obtained in each de-chatter process, and that there is no important tissue structural information lost in a de-chatter image. |
