Topographical analysis in the SEM using an automatic focusing technique

A novel technique for automated topographical analysis in the SEM has been investigated. It utilizes a 16-bit minicomputer arranged to act as an automatic focusing unit. The computer is coupled to the objective lens of the microscope, by means of a digital to analogue converter, and may regulate the excitation of the lens under program control. Further digital-to-analogue converters allow the computer to act as a programmable scan generator by applying ramp waveforms to the scan amplifiers, permitting the beam to be swept over a small sub-region of the field of interest. The video signal is sampled and applied to an analogue-to-digital converter; the resultant binary numbers are stored in computer memory as an array of values representing relative image intensities within a subregion. A differencing algorithm applied to the collected data allows the level of objective lens excitation to be found at which the sharpness of the image is optimized, and the excitation may be related to the working distance for that subregion through a previous calibration experiment. The sensitivity of the method for detecting small height changes is theoretically of the order of 1 μm.

MPC: A paradigm shift for automatic control

Model Predictive Control (MPC) represents a major paradigm shift in the field of automatic control. This radically affects synthesis techniques (illustrated by control of an unstable system) and underlying concepts (illustrated by control of a multivariable system), as well as lifting the control engineer's focus from prescriptions to specifications ('what' not 'how', illustrated by emulation of a conventional autopilot). Part of the objective of this paper is to emphasize the significance of this paradigm shift. Another part is to consider the fact that this shift was missed for many years by the academic community, and what this tells us about teaching and research in the field.

The modified Kanerva model for automatic speech recognition

A parallel processing network derived from Kanerva's associative memory theory Kanerva 1984 is shown to be able to train rapidly on connected speech data and recognize further speech data with a label error rate of 0·68%. This modified Kanerva model can be trained substantially faster than other networks with comparable pattern discrimination properties. Kanerva presented his theory of a self-propagating search in 1984, and showed theoretically that large-scale versions of his model would have powerful pattern matching properties. This paper describes how the design for the modified Kanerva model is derived from Kanerva's original theory. Several designs are tested to discover which form may be implemented fastest while still maintaining versatile recognition performance. A method is developed to deal with the time varying nature of the speech signal by recognizing static patterns together with a fixed quantity of contextual information. In order to recognize speech features in different contexts it is necessary for a network to be able to model disjoint pattern classes. This type of modelling cannot be performed by a single layer of links. Network research was once held back by the inability of single-layer networks to solve this sort of problem, and the lack of a training algorithm for multi-layer networks. Rumelhart, Hinton & Williams 1985 provided one solution by demonstrating the "back propagation" training algorithm for multi-layer networks. A second alternative is used in the modified Kanerva model. A non-linear fixed transformation maps the pattern space into a space of higher dimensionality in which the speech features are linearly separable. A single-layer network may then be used to perform the recognition. The advantage of this solution over the other using multi-layer networks lies in the greater power and speed of the single-layer network training algorithm. © 1989.

Spring-to-Spring Balancing as Energy-Free Adjustment Method in Gravity Equilibrators

Generally, adjustment of gravity equilibrator to a new payload requires energy, e.g. to increase the pre-load of the balancing spring. A novel way of energy-free adjustment of gravity equilibrators is possible by introducing the concept of a storage spring. The storage spring supplies or stores the energy necessary to adjust the balancer spring of the gravity equilibrator. In essence the storage spring mechanism maintains a constant potential energy within the spring mechanism; energy is exchanged between the storage and balancer spring when needed. Various conceptual designs using both zero-free-length springs and regular extension springs are proposed. Two models were manufactured demonstrating the practical embodiments and functionality.

Stereo display with time-multiplexed focal adjustment

In stereo displays, binocular disparity creates a striking impression of depth. However, such displays present focus cues - blur and accommodation - that specify a different depth than disparity, thereby causing a conflict. This conflict causes several problems including misperception of the 3D layout, difficulty fusing binocular images, and visual fatigue. To address these problems, we developed a display that preserves the advantages of conventional stereo displays, while presenting correct or nearly correct focus cues. In our new stereo display each eye views a display through a lens that switches between four focal distances at very high rate. The switches are synchronized to the display, so focal distance and the distance being simulated on the display are consistent or nearly consistent with one another. Focus cues for points in-between the four focal planes are simulated by using a depth-weighted blending technique. We will describe the design of the new display, discuss the retinal images it forms under various conditions, and describe an experiment that illustrates the effectiveness of the display in maximizing visual performance while minimizing visual fatigue. © 2009 SPIE-IS&T.

Automatic object segmentation from calibrated images

This paper addresses the problem of automatically obtaining the object/background segmentation of a rigid 3D object observed in a set of images that have been calibrated for camera pose and intrinsics. Such segmentations can be used to obtain a shape representation of a potentially texture-less object by computing a visual hull. We propose an automatic approach where the object to be segmented is identified by the pose of the cameras instead of user input such as 2D bounding rectangles or brush-strokes. The key behind our method is a pairwise MRF framework that combines (a) foreground/background appearance models, (b) epipolar constraints and (c) weak stereo correspondence into a single segmentation cost function that can be efficiently solved by Graph-cuts. The segmentation thus obtained is further improved using silhouette coherency and then used to update the foreground/background appearance models which are fed into the next Graph-cut computation. These two steps are iterated until segmentation convergences. Our method can automatically provide a 3D surface representation even in texture-less scenes where MVS methods might fail. Furthermore, it confers improved performance in images where the object is not readily separable from the background in colour space, an area that previous segmentation approaches have found challenging. © 2011 IEEE.