Thesis (M.Sc.Eng.) PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you. / The atomic force microscope (AFM) is a prevalent imaging device recognized for its capacity to measure surface topology at the subatomic level. Its exceptional ability to operate in a range of atmospheres from high vacuum to liquid environments and simultaneously quantify various material properties make it particularly well-suited to biological applications. Standard AFMs generate images by transversing a predefined rectangular region with a mechanical probe that maps the surface pixel by pixel. As a result, typical scan times are on the order of seconds to minutes and do not allow for the direct observation of dynamic processes, such as motor protein behavior. State-of-the-art AFMs attempt to improve temporal resolution by employing advanced controllers, converting to mechanical components designed for rapid response, and utilizing scan trajectories that consider actuator dynamics. Successful application of such techniques has delivered scan rates as high as 10 frames per second, with the unfortunate sacrifices of reduced frame size and costly equipment upgrades.
A complementary approach aims to enable the substantial base of commercial AFMs to perform with similar high-speed capabilities by autonomous driving scan trajectories along key features. A previously developed technique, the local raster scan (LRS), follows polymer samples, such as DNA or actin filaments, by detecting structural edges in real time and steering the probe in a sinusoidal path across the strand. While this has been shown to reduce scanning time by one order of magnitude, it is limited by computational complexity and to the imaging of smooth curves.
In this work, we present the local circular scan (LCS), a novel feature-tracking procedure that successfully addresses these restrictions. By utilizing a local reference frame with pragmatically chosen state variables, trajectory calculations are simply reduced to vector operations. Additionally, the self-intersecting, circular trajectory permits more sophisticated filtering, both in real-time and during post-processing.
The contribution of this thesis is the development, implementation, and analysis of the LCS algorithm. A calibration sample with linear, square, and circular features is used for testing. Experimental results demonstrate an ability to track regions of high curvature and robustness to noise. Corrections for sample tilt and thermal drift as well as interpolation techniques used for image processing are detailed. / 2031-01-01
| Identifer | oai:union.ndltd.org:bu.edu/oai:open.bu.edu:2144/21279 |
| Date | January 2014 |
| Creators | Worthey, Jeffrey L. |
| Publisher | Boston University |
| Source Sets | Boston University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
Page generated in 0.0016 seconds