Lester A. Gerhardt

Decoding two-dimensional movement trajectories using electrocorticographic signals in humans

Signals from the brain could provide a non-muscular communication and control system, a brain-computer interface (BCI), for people who are severely paralyzed. A common BCI research strategy begins by decoding kinematic parameters from brain signals recorded during actual arm movement. It has been assumed that these parameters can be derived accurately only from signals recorded by intracortical microelectrodes, but the long-term stability of such electrodes is uncertain. The present study disproves this widespread assumption by showing in humans that kinematic parameters can also be decoded from signals recorded by subdural electrodes on the cortical surface (ECoG) with an accuracy comparable to that achieved in monkey studies using intracortical microelectrodes. A new ECoG feature labeled the local motor potential (LMP) provided the most information about movement. Furthermore, features displayed cosine tuning that has previously been described only for signals recorded within the brain. These results suggest that ECoG could be a more stable and less invasive alternative to intracortical electrodes for BCI systems, and could also prove useful in studies of motor function.

A nonlinear orientation model for global description of fingerprints

A nonlinear orientation model for the representation of the orientation matrix of a fingerprint image is presented. A general procedure to obtain the set of parameters describing this model is shown. Particularly, a piecewise linear orientation function is analysed. The model, compared with a previous zero-pole model, shows substantial decrease in the orientation error. The application of this model for efficient identity verification based on the ridge orientation is explored.

Brain-computer interfaces (BCIs): detection instead of classification.

Many studies over the past two decades have shown that people can use brain signals to convey their intent to a computer through brain-computer interfaces (BCIs). These devices operate by recording signals from the brain and translating these signals into device commands. They can be used by people who are severely paralyzed to communicate without any use of muscle activity. One of the major impediments in translating this novel technology into clinical applications is the current requirement for preliminary analyses to identify the brain signal features best suited for communication. This paper introduces and validates signal detection, which does not require such analysis procedures, as a new concept in BCI signal processing. This detection concept is realized with Gaussian mixture models (GMMs) that are used to model resting brain activity so that any change in relevant brain signals can be detected. It is implemented in a package called SIGFRIED (SIGnal modeling For Real-time Identification and Event Detection). The results indicate that SIGFRIED produces results that are within the range of those achieved using a common analysis strategy that requires preliminary identification of signal features. They indicate that such laborious analysis procedures could be replaced by merely recording brain signals during rest. In summary, this paper demonstrates how SIGFRIED could be used to overcome one of the present impediments to translation of laboratory BCI demonstrations into clinically practical applications.

Real-time detection of event-related brain activity

The complexity and inter-individual variation of brain signals impedes real-time detection of events in raw signals. To convert these complex signals into results that can be readily understood, current approaches usually apply statistical methods to data from known conditions after all data have been collected. The capability to provide meaningful visualization of complex brain signals without the requirement to initially collect data from all conditions would provide a new tool, essentially a new imaging technique, that would open up new avenues for the study of brain function. Here we show that a new analysis approach, called SIGFRIED, can overcome this serious limitation of current methods. SIGFRIED can visualize brain signal changes without requiring prior data collection from all conditions. This capacity is particularly well suited to applications in which comprehensive prior data collection is impossible or impractical, such as intraoperative localization of cortical function or detection of epileptic seizures.

Adaptive integrated manufacturing enterprises: information technology for the next decade

A new vision effecting adaptiveness in integrated manufacturing enterprises for the next decade is formulated. This vision has been developed on the basis of intensive research over the past nine years in Rensselaers industry-sponsored Computer Integrated Manufacturing Program. Built from existing results in both the scientific community and industry, the proposed research agenda calls for new fundamental information technology to enable Adaptive Integrated Manufacturing Enterprises (AIME). It focuses on four major problems: (1) management of multiple systems that operate concurrently over a widely distributed network without a central controller; (2) achievement of an open systems architecture that can accommodate legacy systems as well as add new systems; (3) exploitation of object-oriented technology in production systems with the crucial ability to manage heterogeneous views and propagate changes between views; and (4) modeling of enterprise information requirements for inspection and the utilization of inspection information to create a feedback loop from production to design. These problems and approaches to their solution developed are analyzed. >

Three-dimensional vision

Three-dimensional vision refers to the process of gathering, processing, and interpreting visual data of a three-dimensional environment. A 3-D vision system may employ either passive or active techniques. A passive technique is one that utilizes available light sources (i.e., those necessary to provide general illumination of the viewed environment), whereas an active technique utilizes the projection of a prestructured light pattern to supply “ground truth” information. This chapter describes the calibration of a fixed camera and a fixed laser scanner. Three-dimensional point estimation methods are then discussed. Finally, a working 3-D vision system used for robotic assembly is described.

Non-uniform surface sampling techniques for three-dimensional object inspection

While the uniform sampling method is quite popular for point- wise measurement of manufactured parts, we present three novel sam- pling strategies that emphasize 3D non-uniform inspection capability. They are direct and indirect adaptive sampling and local adjustment sampling. The adaptive sampling strategy is based on a recursive sur- face subdivision process that applies two different approaches. One uses the direct triangular patch subdivision while the other uses the in- direct sectional adaptive approach. The direct adaptive sampling ap- proach can distribute points more closely around edges, corners, and vertices as found on the class of machined products. The indirect adap- tive sampling techniques extend optimum 2D sampling methods to 3D applications. The modified 2D adaptive sampling techniques are used sequentially twice; first, the critical cross sections are optimally selected, and then each section is optimally sampled to develop an accurate geo- metric description using a small number of sampling points. Beyond the practical application value of a technique to inspect curved surface ob- jects, this kind of technique is also of value in understanding the principle of optimum sampling in a 3D sense. The local adjustment sampling strat- egy uses a set of predefined starting points and then finds the local optimum position of each nodal point. This method approximates the object by moving the points toward object edges and corners. The pre- defined starting points sets include uniform and non-uniform sampling distribution generated by the direct adaptive sampling approach. The results show that the initial point sets, when preprocessed by the adap- tive sampling using triangular patches, are moved the shortest distance to edges and corners for global optimum approximation, again showing this methods superiority. The performance comparisons of applying uni- form sampling and adaptive sampling are made based on the MSE mean square error value between the real object surfaces and their approximating surfaces. The adaptive sampling methods exhibit better performance than the uniform sampling methods in reducing the MSE values with fewer sample points. In addition to the performance advan- tage over uniform sampling, the non-uniform sampling techniques we propose also proved to be integratable with view planning for the inspec- tion of products by different manufacturing processes.

Development of a required international experience for undergraduate engineering students

The School of Engineering at Rensselaer has recently announced the REACH Program (Rensselaer Engineering Education Across Cultural Horizons) as a component of our undergraduate engineering degree that we envision will become a requirement for every student. The major feature of this experience will be a bilateral semester abroad with partner international universities, whereby students will spend a full semester studying abroad as a part of their educational program, and students from the partner university will likewise spend a semester at Rensselaer. The Program will begin in 2009 with approximately 25% of the then junior class growing to virtually 100% by 2015. The critical issues in developing and scaling up to such a large program of exchange will be discussed, along with the alternative international experiences that are also under development. We envision that this program will radically alter the culture of our undergraduate programs and our campus environment and will establish Rensselaer as unique among major engineering universities in terms of commitment to providing such an expansive global perspective for all of our students.

The use of laser structured light for 3D surface measurement and inspection

Computer aided inspection (CAI) systems need to be capable of integrating CAD, manufacturing and verification by automating the inspection process and moving it upstream in the manufacturing cycle. This paper discusses structured light systems that are used for obtaining range data of three-dimensional (3D) objects for verification and recognition using a triangulation method. The authors investigate field of view for three configurations of structured light systems and also describe visibility/occlusion for those configurations. Using a simplified camera calibration method, the authors describe how depth information of an object can be obtained using a closed-loop depth estimation method. Because of the system configuration and the properties of laser and camera this sensing method also generates errors. Thus in the paper, measurement errors are also analyzed.<<ETX>>

A class of adaptive model- and object-driven nonuniform sampling methods for 3-D inspection

Inspection of manufactured parts and assemblies often requires large amounts of information in the form of test probe point locations and large amounts of time to perform the inspection. By optimally locating the probe points it is possible to maintain inspection reliability using fewer test probes in a reduced amount of time. We have developed algorithms which use part model and manufacturing process information to generate an optimal probe-point location set for routine inspection in a modelbased, open-loop mode. An alternate set of adaptive algorithms that sequentially generates probe-point locations in an object-based, closed-loop mode characterizes a fault when one is detected. Test results show that the algorithms perform favorably on a large class of surfaces.