On Visual Detection of Highly-occluded Objects for Harvesting Automation in Horticulture

Developing accurate and reliable crop detection algorithms is an important step for harvesting automation in horticulture. This paper presents a novel approach to visual detection of highly-occluded fruits. We use a conditional random field (CRF) on multi-spectral image data (colour and Near-Infrared Reflectance, NIR) to model two classes: crop and background. To describe these two classes, we explore a range of visual-texture features including local binary pattern, histogram of oriented gradients, and learn auto-encoder features. The pro-posed methods are evaluated using hand-labelled images from a dataset captured on a commercial capsicum farm. Experimental results are presented, and performance is evaluated in terms of the Area Under the Curve (AUC) of the precision-recall curves.Our current results achieve a maximum performance of 0.81AUC when combining all of the texture features in conjunction with colour information.

Problems and Algorithms for Sequence Segmentations

The analysis of sequential data is required in many diverse areas such as telecommunications, stock market analysis, and bioinformatics. A basic problem related to the analysis of sequential data is the sequence segmentation problem. A sequence segmentation is a partition of the sequence into a number of non-overlapping segments that cover all data points, such that each segment is as homogeneous as possible. This problem can be solved optimally using a standard dynamic programming algorithm. In the first part of the thesis, we present a new approximation algorithm for the sequence segmentation problem. This algorithm has smaller running time than the optimal dynamic programming algorithm, while it has bounded approximation ratio. The basic idea is to divide the input sequence into subsequences, solve the problem optimally in each subsequence, and then appropriately combine the solutions to the subproblems into one final solution. In the second part of the thesis, we study alternative segmentation models that are devised to better fit the data. More specifically, we focus on clustered segmentations and segmentations with rearrangements. While in the standard segmentation of a multidimensional sequence all dimensions share the same segment boundaries, in a clustered segmentation the multidimensional sequence is segmented in such a way that dimensions are allowed to form clusters. Each cluster of dimensions is then segmented separately. We formally define the problem of clustered segmentations and we experimentally show that segmenting sequences using this segmentation model, leads to solutions with smaller error for the same model cost. Segmentation with rearrangements is a novel variation to the segmentation problem: in addition to partitioning the sequence we also seek to apply a limited amount of reordering, so that the overall representation error is minimized. We formulate the problem of segmentation with rearrangements and we show that it is an NP-hard problem to solve or even to approximate. We devise effective algorithms for the proposed problem, combining ideas from dynamic programming and outlier detection algorithms in sequences. In the final part of the thesis, we discuss the problem of aggregating results of segmentation algorithms on the same set of data points. In this case, we are interested in producing a partitioning of the data that agrees as much as possible with the input partitions. We show that this problem can be solved optimally in polynomial time using dynamic programming. Furthermore, we show that not all data points are candidates for segment boundaries in the optimal solution.

Cooperative Spectrum Sensing Algorithms For OFDM Systems with Frequency Selective Channels

This paper considers the problem of spectrum sensing in cognitive radio networks when the primary user employs Orthogonal Frequency Division Multiplexing (OFDM). We specifically consider the scenario when the channel between the primary and a secondary user is frequency selective. We develop cooperative sequential detection algorithms based on energy detectors. We modify the detectors to mitigate the effects of some common model uncertainties such as timing and frequency offset, IQ-imbalance and uncertainty in noise and transmit power. The performance of the proposed algorithms are studied via simulations. We show that the performance of the energy detector is not affected by the frequency selective channel. We also provide a theoretical analysis for some of our algorithms.

Cyclic Prefix Based Cooperative Sequential Spectrum Sensing Algorithms for OFDM

This paper considers the problem of spectrum sensing in cognitive radio networks when the primary user is using Orthogonal Frequency Division Multiplexing (OFDM). For this we develop cooperative sequential detection algorithms that use the autocorrelation property of cyclic prefix (CP) used in OFDM systems. We study the effect of timing and frequency offset, IQ-imbalance and uncertainty in noise and transmit power. We also modify the detector to mitigate the effects of these impairments. The performance of the proposed algorithms is studied via simulations. We show that sequential detection can significantly improve the performance over a fixed sample size detector.

CUSUM Based Distributed Detection in WSNs

We are concerned with the situation in which a wireless sensor network is deployed in a region, for the purpose of detecting an event occurring at a random time and at a random location. The sensor nodes periodically sample their environment (e.g., for acoustic energy),process the observations (in our case, using a CUSUM-based algorithm) and send a local decision (which is binary in nature) to the fusion centre. The fusion centre collects these local decisions and uses a fusion rule to process the sensors’ local decisions and infer the state of nature, i.e., if an event has occurred or not. Our main contribution is in analyzing two local detection rules in combination with a simple fusion rule. The local detection algorithms are based on the nonparametric CUSUMprocedure from sequential statistics. We also propose two ways to operate the local detectors after an alarm. These alternatives when combined in various ways yield several approaches. Our contribution is to provide analytical techniques to calculate false alarm measures, by the use of which the local detector thresholds can be set. Simulation results are provided to evaluate the accuracy of our analysis. As an illustration we provide a design example. We also use simulations to compare the detection delays incurred in these algorithms.

Joint Self-Iterating Equalization and Detection for Two-Dimensional Intersymbol-Interference Channels

We develop several novel signal detection algorithms for two-dimensional intersymbol-interference channels. The contribution of the paper is two-fold: (1) We extend the one-dimensional maximum a-posteriori (MAP) detection algorithm to operate over multiple rows and columns in an iterative manner. We study the performance vs. complexity trade-offs for various algorithmic options ranging from single row/column non-iterative detection to a multi-row/column iterative scheme and analyze the performance of the algorithm. (2) We develop a self-iterating 2-D linear minimum mean-squared based equalizer by extending the 1-D linear equalizer framework, and present an analysis of the algorithm. The iterative multi-row/column detector and the self-iterating equalizer are further connected together within a turbo framework. We analyze the combined 2-D iterative equalization and detection engine through analysis and simulations. The performance of the overall equalizer and detector is near MAP estimate with tractable complexity, and beats the Marrow Wolf detector by about at least 0.8 dB over certain 2-D ISI channels. The coded performance indicates about 8 dB of significant SNR gain over the uncoded 2-D equalizer-detector system.

Lattice Reduction Aided Detection in Large-MIMO Systems

Lattice reduction (LR) aided detection algorithms are known to achieve the same diversity order as that of maximum-likelihood (ML) detection at low complexity. However, they suffer SNR loss compared to ML performance. The SNR loss is mainly due to imperfect orthogonalization and imperfect nearest neighbor quantization. In this paper, we propose an improved LR-aided (ILR) detection algorithm, where we specifically target to reduce the effects of both imperfect orthogonalization and imperfect nearest neighbor quantization. The proposed ILR detection algorithm is shown to achieve near-ML performance in large-MIMO systems and outperform other LR-aided detection algorithms in the literature. Specifically, the SNR loss incurred by the proposed ILR algorithm compared to ML performance is just 0.1 dB for 4-QAM and < 0.5 dB for 16-QAM in 16 x 16 V-BLAST MIMO system. This performance is superior compared to those of other LR-aided detection algorithms, whose SNR losses are in the 2 dB to 9 dB range.

Generalized Partial Response Equalization and Data-Dependent Noise Predictive Signal Detection Over Media Models for TDMR

Two-dimensional magnetic recording (2-D TDMR) is an emerging technology that aims to achieve areal densities as high as 10 Tb/in(2) using sophisticated 2-D signal-processing algorithms. High areal densities are achieved by reducing the size of a bit to the order of the size of magnetic grains, resulting in severe 2-D intersymbol interference (ISI). Jitter noise due to irregular grain positions on the magnetic medium is more pronounced at these areal densities. Therefore, a viable read-channel architecture for TDMR requires 2-D signal-detection algorithms that can mitigate 2-D ISI and combat noise comprising jitter and electronic components. Partial response maximum likelihood (PRML) detection scheme allows controlled ISI as seen by the detector. With the controlled and reduced span of 2-D ISI, the PRML scheme overcomes practical difficulties such as Nyquist rate signaling required for full response 2-D equalization. As in the case of 1-D magnetic recording, jitter noise can be handled using a data-dependent noise-prediction (DDNP) filter bank within a 2-D signal-detection engine. The contributions of this paper are threefold: 1) we empirically study the jitter noise characteristics in TDMR as a function of grain density using a Voronoi-based granular media model; 2) we develop a 2-D DDNP algorithm to handle the media noise seen in TDMR; and 3) we also develop techniques to design 2-D separable and nonseparable targets for generalized partial response equalization for TDMR. This can be used along with a 2-D signal-detection algorithm. The DDNP algorithm is observed to give a 2.5 dB gain in SNR over uncoded data compared with the noise predictive maximum likelihood detection for the same choice of channel model parameters to achieve a channel bit density of 1.3 Tb/in(2) with media grain center-to-center distance of 10 nm. The DDNP algorithm is observed to give similar to 10% gain in areal density near 5 grains/bit. The proposed signal-processing framework can broadly scale to various TDMR realizations and areal density points.

FPGA-based Tabu Search for Detection in Large-Scale MIMO Systems

The increasing scale of Multiple-Input Multiple- Output (MIMO) topologies employed in forthcoming wireless communications standards presents a substantial implementation challenge to designers of embedded baseband signal processing architectures for MIMO transceivers. Specifically the increased scale of such systems has a substantial impact on the perfor- mance/cost balance of detection algorithms for these systems. Whilst in small-scale systems Sphere Decoding (SD) algorithms offer the best quasi-ML performance/cost balance, in larger systems heuristic detectors, such Tabu-Search (TS) detectors are superior. This paper addresses a dearth of research in architectures for TS-based MIMO detection, presenting the first known realisations of TS detectors for 4 × 4 and 10 × 10 MIMO systems. To the best of the authors’ knowledge, these are the largest single-chip detectors on record.

Face Detection and Facial Feature Localization for multi-pose faces and complex backgroundimages

The objective of this thesis work, is to propose an algorithm to detect the faces in a digital image with complex background. A lot of work has already been done in the area of face detection, but drawback of some face detection algorithms is the lack of ability to detect faces with closed eyes and open mouth. Thus facial features form an important basis for detection. The current thesis work focuses on detection of faces based on facial objects. The procedure is composed of three different phases: segmentation phase, filtering phase and localization phase. In segmentation phase, the algorithm utilizes color segmentation to isolate human skin color based on its chrominance properties. In filtering phase, Minkowski addition based object removal (Morphological operations) has been used to remove the non-skin regions. In the last phase, Image Processing and Computer Vision methods have been used to find the existence of facial components in the skin regions.This method is effective on detecting a face region with closed eyes, open mouth and a half profile face. The experiment’s results demonstrated that the detection accuracy is around 85.4% and the detection speed is faster when compared to neural network method and other techniques.

Automatic system for thermal damage detection in manufacturing process with Internet monitoring

This work involved the development of a smart system dedicated to surface burning detection in the grinding process through constant monitoring of the process by acoustic emission and electrical power signals. A program in Visual Basic® for Windows® was developed, which collects the signals through an analog-digital converter and further processes them using burning detection algorithms already known. Three other parameters are proposed here and a comparative study carried out. When burning occurs, the newly developed software program sends a control signal warning the operator or interrupting the process, and delivers process information via the Internet. Parallel to this, the user can also interfere in the process via Internet, changing parameters and/or monitoring the grinding process. The findings of a comparative study of the various parameters are also discussed here. Copyright © 2006 by ABCM.

Intrusion detection system using optimum-path forest

Intrusion detection systems that make use of artificial intelligence techniques in order to improve effectiveness have been actively pursued in the last decade. Neural networks and Support Vector Machines have been also extensively applied to this task. However, their complexity to learn new attacks has become very expensive, making them inviable for a real time retraining. In this research, we introduce a new pattern classifier named Optimum-Path Forest (OPF) to this task, which has demonstrated to be similar to the state-of-the-art pattern recognition techniques, but extremely more efficient for training patterns. Experiments on public datasets showed that OPF classifier may be a suitable tool to detect intrusions on computer networks, as well as allow the algorithm to learn new attacks faster than the other techniques. © 2011 IEEE.

Detection and Tracking of Dynamic Objects. A MultiRobot approach to Critical Infrastructures Surveillance.

Despite that Critical Infrastructures (CIs) security and surveillance are a growing concern for many countries and companies, Multi Robot Systems (MRSs) have not been yet broadly used in this type of facilities. This dissertation presents a novel study of the challenges arisen by the implementation of this type of systems and proposes solutions to specific problems. First, a comprehensive analysis of different types of CIs has been carried out, emphasizing the influence of the different characteristics of the facilities in the design of a security and surveillance MRS. One of the most important needs for the surveillance of a CI is the detection of intruders. From a technical point of view this problem can be abstracted as equivalent to the Detection and Tracking of Mobile Objects (DATMO). This dissertation proposes algorithms to solve this specific problem in a CI environment. Using 3D range images of the environment as input data, two detection algorithms for ground robots have been developed. These detection algorithms provide a list of moving objects in the robot detection area. Direct image differentiation and computer vision techniques are used when the robot is static. Alternatively, multi-layer ground reconstructions are compared to detect the dynamic objects when the robot is moving. Since CIs usually spread over large areas, it is very useful to incorporate aerial vehicles in the surveillance MRS. Therefore, a moving object detection algorithm for aerial vehicles has been also developed. This algorithm compares the real optical flow obtained from a down-face oriented camera with an artificial optical flow computed using a RANSAC based homography matrix. Two tracking algorithms have been developed to follow the moving objects trajectories. These algorithms can efficiently handle occlusions and crossings, as well as exchange information among robots. The multirobot tracking can be applied to any type of communication structure: centralized, decentralized or a combination of both. Even more, the developed tracking algorithms are independent of the detection algorithms and could be potentially used with other detection procedures or even with static sensors, such as cameras. In addition, using the 3D point clouds available to the robots, a relative localization algorithm has been developed to improve the position estimation of a given robot with observations from other robots. All the developed algorithms have been extensively tested in different simulated CIs using the Webots robotics simulator. Furthermore, the algorithms have also been validated with real robots operating in real scenarios. In conclusion, this dissertation presents a multirobot approach to Critical Infrastructure Surveillance, mainly focusing on Detecting and Tracking Dynamic Objects.

Marine mammal species detection and classification

Thesis (Ph.D.)--University of Washington, 2016-04

GANN: Genetic algorithm neural networks for the detection of conserved combinations of features in DNA

Background: The multitude of motif detection algorithms developed to date have largely focused on the detection of patterns in primary sequence. Since sequence-dependent DNA structure and flexibility may also play a role in protein-DNA interactions, the simultaneous exploration of sequence-and structure-based hypotheses about the composition of binding sites and the ordering of features in a regulatory region should be considered as well. The consideration of structural features requires the development of new detection tools that can deal with data types other than primary sequence. Results: GANN ( available at http://bioinformatics.org.au/gann) is a machine learning tool for the detection of conserved features in DNA. The software suite contains programs to extract different regions of genomic DNA from flat files and convert these sequences to indices that reflect sequence and structural composition or the presence of specific protein binding sites. The machine learning component allows the classification of different types of sequences based on subsamples of these indices, and can identify the best combinations of indices and machine learning architecture for sequence discrimination. Another key feature of GANN is the replicated splitting of data into training and test sets, and the implementation of negative controls. In validation experiments, GANN successfully merged important sequence and structural features to yield good predictive models for synthetic and real regulatory regions. Conclusion: GANN is a flexible tool that can search through large sets of sequence and structural feature combinations to identify those that best characterize a set of sequences.