Scanning Tunneling Microscopy (STM) is commonly used to obtain images of surfaces with unparalleled resolution. However since the early days of the development of this technique and home-built microscopes, STM field has become a veritable industry churning out hundreds of images. Studying nanostructures is a particularly exciting application of STM, but often it is desirable to extract quantitative data (e.g. terrace width distributions, average feature sizes, etc.) in addition to obtaining a clear and pleasant-looking image.

Moreover, to obtain statistically significant results often tens or hundreds of images need to be analyzed, making it prohibitively time-consuming and inaccurate to do the analysis "by hand". Automated procedures are desirable for processing large volumes of data, and I have developed a couple of such procedures.

Step and Wire Edge Detection
Island Aspect Ratio Distributions

Both of the above procedures use some simple image analysis algorithms such as edge detection, but also feature some more advanced tools like Principal Component Analysis (PCA) and neural-network classifiers.

Nanowiz Main Page