Embedded systems are dependent on low-power, miniaturized instrumentation. Comparator circuits are
common elements in applications for digital threshold detection. A multi-level, memory-based logic approach is in
development that offers potential benefits in power usage and size with respect to traditional binary logic systems.
Basic 4-bit operations with CMOS gates and comparators are chosen to compare circuit implementations of binary
structures and quaternary equivalents. Circuit layouts and functional operation are presented. In particular, power
characteristics and transistor count are examined. The potential for improved embedded systems based on the multilevel,
memory-based logic is discussed.
Biomedical journal articles contain a variety of image types that can be broadly classified into two categories: regular
images, and graphical images. Graphical images can be further classified into four classes: diagrams, statistical figures,
flow charts, and tables. Automatic figure type identification is an important step toward improved multimodal (text +
image) information retrieval and clinical decision support applications. This paper describes a feature-based learning
approach to automatically identify these four graphical figure types. We apply Evolutionary Algorithm (EA), Binary
Particle Swarm Optimization (BPSO) and a hybrid of EA and BPSO (EABPSO) methods to select an optimal subset of
extracted image features that are then classified using a Support Vector Machine (SVM) classifier. Evaluation performed
on 1038 figure images extracted from ten BioMedCentral® journals with the features selected by EABPSO yielded
classification accuracy as high as 87.5%.
Biomedical images are often referenced for clinical decision support (CDS), educational purposes, and research. The
task of automatically finding the images in a scientific article that are most useful for the purpose of determining
relevance to a clinical situation is traditionally done using text and is quite challenging. We propose to improve this
by associating image features from the entire image and from relevant regions of interest with biomedical concepts
described in the figure caption or discussion in the article. However, images used in scientific article figures are
often composed of multiple panels where each sub-figure (panel) is referenced in the caption using alphanumeric
labels, e.g. Figure 1(a), 2(c), etc. It is necessary to separate individual panels from a multi-panel figure as a first step
toward automatic annotation of images.
In this work we present methods that add make robust our previous efforts reported here. Specifically, we address
the limitation in segmenting figures that do not exhibit explicit inter-panel boundaries, e.g. illustrations, graphs, and
charts. We present a novel hybrid clustering algorithm based on particle swarm optimization (PSO) with fuzzy logic
controller (FLC) to locate related figure components in such images.
Results from our evaluation are very promising with 93.64% panel detection accuracy for regular (non-illustration)
figure images and 92.1% accuracy for illustration images. A computational complexity analysis also shows that PSO
is an optimal approach with relatively low computation time. The accuracy of separating these two type images is
98.11% and is achieved using decision tree.
A computer-vision monitoring system is demonstrated that automatically detects the presence and location of people.
The approach investigated the potential for real-time, automated surveillance and tracking in a realistic environment.
Economy was obtained by the use of gray-scale, fixed perspective images and efficiency was obtained by the use of
selected object features and a neural-network-processing algorithm. The system was applied to pedestrian traffic on an
outdoor bridge and consequently had to handle complex images. Image sequences of single and multiple people were
used with differences in clothing, position, lighting, season, etc. A two-stage algorithm was implemented in which (1)
new objects were identified in a highly variable scene and (2) the objects were classified with a back-propagation neural
network. The image processing techniques included segmentation and filtering and the neural network used fourteen
object features as inputs. The implementation had excellent people-discrimination accuracy despite the noise in the
images and had low computational complexity with respect to alternative techniques.
KEYWORDS: Sensors, Land mines, Metals, Wavelets, General packet radio service, Mining, Detection and tracking algorithms, Electromagnetic coupling, Robotics, Neural networks
This paper presents some advancement in the detection algorithms using EMI sensor, GPR sensor and their fusion. In the EMI algorithm, we propose the application of the weighted distributed density (WDD) functions on the wavelet domain and the time domain of the EMI data for feature based detection. A multilayer perceptron technique is then applied to discriminate between mine and clutter objects based on the wavelet domain and time domain features separately. When the results from the two domains are fused together, the probability of false alarms is reduced by a factor of two. The enhancement in the GPR algorithm includes the depth processing which selects a certain data segment below the ground surface for detection, as well as utilizing the phase variation of the signal return across a mine to achieve better detection. Finally, we present fusion results from EMI and GPR sensors to demonstrate that the two sensors provide complementary information and when they are properly fused together the probability of false alarm can be reduced significantly.
KEYWORDS: General packet radio service, Mining, Electromagnetic coupling, Sensors, Algorithm development, Land mines, Ground penetrating radar, Electromagnetism, Radar, Data conversion
An analysis of the utility of region-based processing of Ground Penetrating Radar (GPR) and Electromagnetic Induction (EMI) is presented. Algorithms for re-sampling GPR data acquired over non-rectangular and non-regular grids are presented. Depth-dependent whitening is used to form GPR images as functions of depth bins. Shape, size, and contrast-based features are used to distinguish mines from non-mines. The processing is compared to point-wise processing of the same data. Comparisons are made to GPR data collected by machine and by humans. Evaluations are performed on calibration data, for which the ground truth is known to the algorithm developers, and blind data, for which the ground truth is not known to the algorithm developers.
Landmine detection using metal detector (MD) and ground penetrating radar (GPR) sensors in hand-held units is a difficult problem. Detection difficulties arise due to: 1) the varying composition and type of metal in landmines, 2) the time-varying nature of background and 3) the variation in height and velocity of the hand-held unit in data measurement. In prior research, spatially distributed MD features were explored for differentiating landmine signatures from background and non-landmine objects. These features were computed based on correlating sequences of MD energy values with six weighted density distribution functions. In this research the effectiveness of these features to detect landmines of varying metal composition and type is investigated. Experimental results are presented from statistical analysis for feature assessment. Preliminary experimental results are also presented for evaluating the impact on MD feature calculations from varying height and sweep rate of the hand-held unit for data acquisition.
Multi-band medium wave infrared (MWIR) image data collected from the Lightweight Airborne multispectral Minefield Detection-Interim (LAMD-I) program is examined for the detection of surface landmines. Because the orientation of the image acquisition from aircraft with respect to the mine and the minefield is unknown, there is a need to develop an orientation invariant-based approach for landmine and minefield detection. A rotation invariant circular harmonics transform (CHT)-based approach is presented for surface landmine detection. The magnitude information from the CHT is used for finding mine-like regions within the MWIR images. A three-tiered hierarchical thresholding technique provides the basis for highlighting potential surface landmines. Mine shape and size information are used for generating landmine confidence values. Surface landmine detection capability is presented for 82 MWIR broadband images with sand and short and long grass terrain conditions for daytime and nighttime acquired MWIR image data. Receiver operator characteristic (ROC) curves are used for comparing experimental results from this technique with an existing an adaptive multi-band CFAR detector (RX approach).
Microarray technology is increasingly used as a means of high throughput analysis of human, non-human and plant genomes. Manual methods of array production have inherent imperfections and variations of the quality of output data derived from these arrays. For paired microarray images acquired using manual methods of array production, image registration is necessary for aligning corresponding spots for comparison. In this research, a dynamic programming technique is investigated for registering and comparing microarray image pairs. The output of the cost function developed provides a similarity measure between the images and can be applied to evaluating the quality of the image pair. The backward solution from the dynamic programming algorithm developed provides the basis for registering the image pairs. The image registration technique provides for an optimal alignment of the image pairs. The aligned images facilitate spot-to-spot comparisons between the image pairs for detecting specific genetic expressions that could be related to bio-medical functions.
Sensor fusion issues in a streamlined assimilation of multi-sensor information for landmine detection are discussed. In particular multi-sensor fusion in hand-held landmine detection system with ground penetrating radar (GPR) and metal detector sensors is investigated. The fusion architecture consists of feature extraction for individual sensors followed by a feed-forward neural network training to learn the feature space representation of the mine/no-mine classification. A correlation feature from GPR, and slope and energy feature from metal detector are used for discrimination. Various fusion strategies are discussed and results compared against each other and against individual sensors using ROC curves for the available multi-sensor data. Both feature level and decision level fusion have been investigated. Simple decision level fusion scheme based on Dempster-Shafer evidence accumulation, soft AND, MIN and MAX are compared. Feature level fusion using neural network training is shown to provide best results. However comparable performance is achieved using decision level sensor fusion based on Dempster-Shafer accumulation. It is noted that, the above simple feed-forward fusion scheme lacks a means to verify detections after a decision has been made. New detection algorithms that are more than anomaly detectors are needed. Preliminary results with features based on independent component analysis (ICA) show promising results towards this end.
KEYWORDS: Land mines, Metals, Neural networks, Mining, Sensors, General packet radio service, Digital filtering, Electromagnetic coupling, Computer engineering, Palladium
Land mine detection using metal detector (MD) and ground penetrating radar (GPR) sensors in hand-held units is a difficult problem. Detection difficulties arise due to: 1) the varying composition and type of metal in land mines, 2) the time-varying nature of background and 3) the variation in height and velocity of the hand-held unit in data measurement. This research introduces new spatially distributed MD features for differentiating land mine signatures from background. The spatially distributed features involve correlating sequences of MD energy values with six weighted density distribution functions. These features are evaluated using a standard back propagation neural network on real data sets containing more than 2,300 mine encounters of different size, shape, content and metal composition that are measured under different soil conditions.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.