<p> Simulation can be an invaluable tool, particularly in fields such as robotics where physical experiments can be extremely expensive, time consuming, and even dangerous. However, the value of a simulator is directly related to its ability to accurately and reliably model physical phenomena such as intermittent contact. In virtually all multibody simulators available today, the standard methods of contact identification and response treat the free space between pairs of bodies as a convex set, when it is in fact non-convex. To reconcile, simulators typically use very small time steps and employ numerous <i> ad hoc</i> corrections, many of which have become commonplace and include allowing interpenetration to occur, arbitrarily limiting response forces, misrepresenting body geometries, and "freezing" bodies with relatively low velocities. Indeed, a vast body of literature exists addressing the many symptoms of dynamic instability due to poor contact determination. </p><p> I herein present a formulation of non-penetration constraints between pairs of bodies which accounts for all possible combinations of active contact. This is the first formulation that accurately models the body geometries near points of <i>potential</i> contact, simultaneously preventing interpenetration while allowing bodies to traverse accurately through the surrounding free space. Unlike the standard approach, this method does not need to guess at which contacts to enforce. This new formulation is easy to incorporate into existing simulation methods, improves accuracy by many orders of magnitude, and is stable for even large time steps. </p><p> Additionally, I present the RPI-MATLAB-Simulator (RPIsim), an open source tool for efficient research and practical teaching in multibody dynamics. RPIsim is designed to be easy to use and easily extended. Students being introduced to dynamics for the first time have no problem creating and running simulations, even with a limited programming background. Researchers can utilize the existing code base to support work on specific areas since it is easy to replace or extend individual modules of the simulator.</p>

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2008. / Source: Dissertation Abstracts International, Volume: 69-05, Section: B, page: 3085. Adviser: Steven M. LaValle. Includes bibliographical references (leaves 125-136) Available on microfilm from Pro Quest Information and Learning.

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2007. / Source: Dissertation Abstracts International, Volume: 68-11, Section: B, page: 7445. Adviser: Steven M. LaValle. Includes bibliographical references (leaves 134-145) Available on microfilm from Pro Quest Information and Learning.

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2008. / Source: Dissertation Abstracts International, Volume: 69-11, Section: B, page: 6944. Adviser: Steven M. LaValle. Includes bibliographical references (leaves 98-109) Available on microfilm from Pro Quest Information and Learning.

<p> Field mapping and estimation become a challenging problem, with their various applications on non-linear estimation, geolocation, and positioning systems. In this research, we develop novel algorithms based on interval analysis and introduce a solution for autonomous map construction, field mapping, geolocation, and simultaneous localization and mapping (SLAM), providing applications on indoor geolocation and other potential areas. </p><p> Generally, the localization algorithm includes a quasi-state estimation and a dynamic estimation. Quasi-static estimation collects each single measurement and give a group of estimation intervals on the pre-constructed field map. Results from quasi-static estimation are processed into the dynamic estimation algorithm, having properties of removing redundant intervals while keep the best estimation results. Sizes of estimation from quasi-static estimation are proved to be related to the resolution of the map and the quality of the sensor. Based on quasi-state estimation algorithm, we develop an algorithm to fuse different type of measurements and discuss the condition when this algorithm an be applied effectively. </p><p> Having theoretical guarantees, we apply these algorithms to augment the accuracy of cell phone geolocation by taking advantage of local variations of magnetic intensity. Thus, the sources of disturbances to magnetometer readings caused indoors are effectively used as beacons for localization. We construct a magnetic intensity map for an indoor environment by collecting magnetic field data over each floor tile. We then test the algorithms without position initialization and obtain indoor geolocation to within 2m while slowly walking over a complex path of 80 meters. The geolocation errors are smaller in the vicinity of large magnetic disturbances. After fusing the magnetometer measurement with inertial measurements on the cell phone, the algorithm yields even smaller geolocation errors of under 50cm for a moving user. </p><p> The map construction and geolocation algorithms are then extended to realize the SLAM, with hierarchical structure of estimation update and localization update. When a new user steps into a random map, the dead reckoning algorithm with assistance of IMU and Kalman filter provides initial estimation of position on the map, which coordinates the corresponding reading of magnetic field intensity as well as all other sources such as WiFi received signal strength (RSS), to construct an initial map. Based on the initial map, we then apply the localization algorithm to estimate new geolocations consequently and fuse the estimation intervals both from IMU and from crowd-sourced field maps to reduce the estimation size and eventually revise the map as well as the geolocation. </p><p> In this research, we have built up mathematical model and developed mathematical solutions with corresponding theories and proofs. Our theoretical results connect geolocation accuracy to combinations of sensor and map properties. </p>

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2008. / Source: Dissertation Abstracts International, Volume: 69-05, Section: B, page: 3247. Adviser: Seth Hutchinson. Includes bibliographical references (leaves 132-136) Available on microfilm from Pro Quest Information and Learning.

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2008. / Source: Dissertation Abstracts International, Volume: 69-11, Section: B, page: 6842. Adviser: Stephanie Alexander. Includes bibliographical references (leaves 83-85) Available on microfilm from Pro Quest Information and Learning.

<p> In this work, a framework for a system which will intelligently assign an edge detection filter to an image based on features taken from the image is introduced. The framework has four parts: the learning stage, image feature extraction, training filter creation, and filter selection training. Two prototypes systems of this framework are given. The learning stage for these systems is the Berkeley Segmentation Database coupled with the Baddelay Delta Metric. Feature extraction is performed using a GIST methodology which extracts color, intensity, and orientation information. The set of image features are used as the input to a single hidden layer feed forward neural network trained using back propagation. The system trains against a set of linear cellular automata filters which are determined to best solve the <i> edge image</i> according to the Baddelay Delta Metric. One system uses cellular automata augmented with a fuzzy rule. The systems are trained and tested against the images from the Berkeley Segmentation Database. The results from the testing indicate that systems built on this framework can perform better than standard methods of edge detection on average across many types of images.</p>

<p>Time-lapse live cell imaging has been increasingly employed by biological and biomedical researchers to understand the underlying mechanisms in cell physiology and development by investigating behavior of cells. This trend has led to a huge amount of image data, the analysis of which becomes a bottleneck in related research. Consequently, how to efficiently analyze the data is emerging as one of the major challenges in the fields. </p><p> Computer vision analysis of non-fluorescent microscopy images, representatively phase-contrast microscopy images, promises to realize a long-term monitoring of live cell behavior with minimal perturbation and human intervention. To take a step forward to such a system, this thesis proposes computer vision algorithms that monitor cell growth, migration, and differentiation by detecting three cellular events&mdash;mitosis (cell division), apoptosis (programmed cell death), and differentiation&mdash;and tracking individual cells. Among the cellular events, to the best our knowledge, apoptosis and a certain type of differentiation, namely muscle myotubes, have never been detected without fluorescent labeling. We address these challenging problems by developing computer vision algorithms adopting phase contrast microscopy. We also significantly improve the accuracy of mitosis detection and cell tracking in phase contrast microscopy over previous methods, particularly under non-trivial conditions, such as high cell density or confluence. We demonstrate the usefulness of our methods in biological research by analyzing cell behavior in scratch wound healing assays. The automated system that we are pursuing would lead to a new paradigm of biological research by enabling quantitative and individualized assessment in behavior of a large population of intact cells. </p>

<p> A software prototype for autonomous 3D scanning of uncooperatively rotating orbital debris using a point cloud sensor is designed and tested. The software successfully generated 3D models under conditions that simulate some on-orbit orbit challenges including relative motion between observer and target, inconsistent target visibility and a target with more than one plane of symmetry. The model scanning software performed well against an irregular object with one plane of symmetry but was weak against objects with 2 planes of symmetry. </p><p> The suitability of point cloud sensors and algorithms for space is examined. Terrestrial Graph SLAM is adapted for an uncooperatively rotating orbital debris scanning scenario. A joint EKF attitude estimate and shape similiarity loop closure heuristic for orbital debris is derived and experimentally tested. The binary Extended Fast Point Feature Histogram (EFPFH) is defined and analyzed as a binary quantization of the floating point EFPFH. Both the binary and floating point EPFH are experimentally tested and compared as part of the joint loop closure heuristic.</p>