One of the great diffculties of carrying out Terahertz experiments is the alignment of the laser beam that is
used to generate the Terahertz radiation. High precision and stability of the beam direction is required over
long time periods. Manual adjustments must be made regularly due to the undesired fluctuations of the laser
beam direction. A dynamic laser beam alignment system actively reduces misalignment from vibration sources,
thermal gradients and mechanical creep. Commercial automatic systems for laser beam stabilization exist but are
expensive especially when they must be employed in multiples. We have designed a system that is straightforward
and inexpensive using a simple mirror actuation device based on the concept of thermal expansion. Nichrome
wire coils surrounding four aluminium rods are used in the mirror actuator. By varying the current through
these coils the amount of expansion caused by the dissipated heat in the rods can be controlled. Our circuit
consists of a number of stages including a novel feedback design that prevents the rod from ever reaching its
maximum length. Our hardware implementation is able to automatically compensate laser beam pointing in
real-time and over extended periods of time.
The design of a motion detector as a robust velocity estimator for real-world applications is massively challenging. Apart from accuracy and reliability it is difficult to achieve operation in real-time. The design should also be small, low cost, low powered and easily integratible. The Reichardt Correlator is often chosen for velocity
estimation due to its accuracy. Unfortunately, the Reichardt Correlator is not a robust estimator of velocity as its output depends on specific aspects of stimuli such as the brightness and spatial frequency. In the literature, robustness is usually achieved through a highly elaborated Reichardt Correlator. Many of these elaborations are difficult to implement in VLSI and they impact significantly on the ability of the motion detector to operate in
real-time. Our simple hardware approach is an analog/digital hybrid in VLSI that utilizes a simple front-end design for the pre-processing of input and does not require a specialised VLSI process. The minimized analog parts improve the reliability and low power consumption of the system. The proposed digital part is simple,
compact and technology independent. Simulations verify that this design is capable of significantly reducing the dependency of the motion detection model on image brightness.
Insects with their amazing visual system are able to perform exceptional navigational feats. In order to
understand how they perform motion detection and velocity estimation, much work has been done in the past
40 years and many models of motion detection have been proposed. One of the earliest and most prominent
models is the Reichardt correlator model. We have elaborated the Reichardt correlator model to include
additional non-linearities that mimic known properties of the insect motion pathway, including logarithmic
encoding of luminance and saturation at various stages of processing. In this paper, we compare the response
of our elaborated model with recordings from fly HS neurons to naturalistic image panoramas. Such responses
are dominated by noise which is largely non-random. Deviations in the correlator response are likely due to
the structure of the visual scene, which we term Pattern noise. Pattern noise is investigated by implementing
saturation at different stages in our model and comparison of each of these models with the physiological data
from the fly is performed using cross covariance technique.
Proc. SPIE. 6414, Smart Structures, Devices, and Systems III
KEYWORDS: Digital signal processing, Detection and tracking algorithms, Data modeling, Spatial frequencies, Sensors, Motion detection, Very large scale integration, Analog electronics, Motion models, Optical correlators
A small low-cost motion detector would have widespread applications in visual control systems such as miniature
unmanned aerial vehicles and collision avoidance systems. In the last 20 years a number of analog VLSI chips have
been developed which incorporate both photodetection and motion computation on the same chip. Nevertheless,
artificial real-time vision and simple seeing systems remain a massive challenge mainly because the environment
greatly impacts on their performance. On the other hand, biological systems have, through years of evolution,
come up with a number of simple but clever solutions. The Reichardt Correlator is a biologically inspired model
for motion detection. However, the basic model is not a robust estimator of velocity. The accuracy and reliability
of this model can be significantly improved through various elaborations. VLSI is ideally suited to the parallel
processing seen in nature because it allows for high device integration density and complex implementation
of complex functions. Howsoever, VLSI poses some serious bounds on the types of elaborations that can be
implemented. We have explored this problem and will present a number of improved models with robust outputs
that are practical in terms of real time implementation in microchips.
The early detection of cancers is critical with respect to treatment and patient survival. Biopsy techniques that
are currently employed for such diagnoses are invasive, time consuming and costly. A Terahertz (THz) imaging
system potentially provides a fast and non-invasive way to detect and diagnose cancer. While there is proof of
concept that THz can distinguish cancerous and normal tissue, the mechanisms underlying this differentiation
are not well understood. A better understanding of THz spectral data can be gained through computational
pattern recognition and related multivariate statistical tools. These allow for the differentiation of data into
discrete and disjoint groups. Such separation of THz spectral data can provide complex information about
diseased tissue, which can be used as a tool for distinguishing cancerous from non-cancerous cells as well as,
discriminating between cancers at various developmental stages and, between different types of cancer.
In this paper we compare the value of different molecular modeling techniques for the prediction of vibrational
modes, especially in the mid- and far-infrared region. There is a wide range of different levels of theory available
for molecular modelling - the choice depending on the kind of system to be investigated. For our calculations
we use different theoretical approaches such as Hartree-Fock and Density functional theory. We also compare
the performances of two available electronic structure programs-Gamess-US and Gaussian03. As examples,
we use two different retinoids - all-trans retinal and all-trans retinoic acid - derivatives of Vitamin A.
Flying insects are able to manoeuvre through complex environments with remarkable ease and accuracy despite their simple visual system. Physiological evidence suggests that flight control is primarily guided by a small system of neurons tuned to very specific types of complex motion. This system is a promising model for bio-inspired approaches to low-cost artificial motion analysis systems, such as collision avoidance devices. A number of models of motion detection have been proposed, with the basic model being the Reichardt Correlator. Electrophysiological data suggest a variety of non-linear elaborations, which include compressive non-linearities and adaptive feedback of local motion detector outputs. In this paper we review a number of computational models for motion detection from the point of view of ease of implementation in low cost VLSI technology. We summarise the features of biological motion analysis systems that are important for the design of real-time artificial motion analysis systems. Then we report on recent progress in bio-inspired analog VLSI chips that capture properties of biological neural computation.
We investigate the classification of the T-ray response of normal human bone cells and human osteosarcoma cells, grown in culture. Given the magnitude and phase responses within a reliable spectral range as features for input vectors, a trained support vector machine can correctly classify the two cell types to some extent. Performance of the support vector machine is deteriorated by the curse of dimensionality, resulting from the comparatively large number of features in the input vectors. Feature subset selection methods are used to select only an optimal number of relevant features for inputs. As a result, an improvement in generalization performance is attainable, and the selected frequencies can be used for further describing different mechanisms of the cells, responding to T-rays. We demonstrate a consistent classification accuracy of 89.6%, while the only one fifth of the original features are retained in the data set.
Terahertz spectroscopy, which investigates the electromagnetic spectrum of samples between 0.1 and 10 THz, allows not only for exploration of molecular structures but also of molecular dynamics. One difficulty in performing THz spectroscopy is that the data can be noisy and difficult to interpret. Ab initio molecular modelling has recently become more and more useful in the prediction of, for example, molecular structures, dynamic states and isomeric forms. Since the structure of biomolecules is closely related to their functionality there are broad ranging applications in biomedicine, for example in DNA sensing. An a priori knowledge of the expected THz spectra allows for improved experimentation. There is a growing and recognised need for THz spectroscopic databases to be created and made available along with classifiers that are able to effectively
detect a specific substance. We show, for a specific example, the 9-cis and all-trans retinal isomers, how ab initio molecular orbital calculations and quantum chemical modelling programs, such as Gamess, can aid in this endeavour.
A simple method to extract the far-infrared dielectric parameters of a homogeneous material from terahertz signals is explored in this paper. Provided with a reference, sample-probing terahertz signal and a known sample thickness, the method can determine the underlying complex refractive index of the sample within a few iterations based on the technique of fixed-point iteration. The iterative process is guaranteed to converge and gives the correct parameters when the material thickness exceeds 200 μm at a frequency of 0.1 THz or 20 μm at a frequency of 1.0 THz.
Terahertz imaging is presently in its exploratory stage. Although plots of time versus terahertz amplitude, and frequency versus terahertz magnitude are some of the most common ways of analyzing terahertz data, no standard rendering technique has been established. While existing methods are indispensable, improvements to how terahertz data is rendered and analyzed should be explored so that new techniques can complement existing ones and/or provide a means of displaying new information that existing methods cannot. This paper reports on one solution to terahertz imaging: an implementation of a new form of phase contrast imaging, which is based on a well-established technique for optical microscopy. This will provide us with a further way of interpreting information from terahertz imaging systems.
Terahertz wavelengths can pass through dry, non-polar, non-metallic materials that are opaque at visible wavelengths. Moreover they can be manipulated using millimeter wave and quasi-optical techniques to form an image. Sensing in this band potentially provides advantages in a number of areas of interest for security and defense, such as screening of personnel for hidden objects, and the detection of chemical and biological agents. This paper reviews recent research into THz applications by groups across Europe, the US, Australasia, and the UK. Several private companies are developing smaller and cheaper reliable devices allowing for commercialisation of these applications. While there are a number of challenges to be overcome there is little doubt that THz technologies will play a major role in the near future for advancement of security, public health and defense.
Insects have a very efficient visual system that helps them to
perform extraordinarily complicated navigational acts and
precisely controlled aerobatic flight. Physiological evidence
suggests that flight control is guided by a small system of
'tangential' neurons tuned to very specific types of complex
motion by the way that they collate information from local motion
detectors. One class of tangential neurons, the 'horizontal
system' (HS) neurons, respond with opponent graded responses to
yaw stimuli. Using the results of physiological experiments, we
have developed a model, based on an array of Reichardt correlators, for the receptive field of HS neurons that view optical flow along the equator. Our model incorporates additional non-linearities that mimic known properties of the insect motion pathway, including logarithmic encoding of luminance, saturation and motion adaptation (adaptive gain-control). In this paper, we compare the response of our elaborated model with fly HS neuron responses to naturalistic image panoramas. Such responses are dominated by noise which is largely non-random. Deviations in the correlator response are likely due to the structure of the visual scene, which we term "Pattern noise". To investigate the influence of anisotropic features in producing pattern noise, we presented a panoramic image at various initial positions, and versions of the same image modified to disrupt vertical contours. We conclude that the response of the fly neurons shows evidence of local saturation at key stages in the motion pathway. This saturation reduces the effect of pattern noise and improves the coding of velocity. Our model provides an excellent basis for the development of biomimetic yaw sensors for robotic applications.