Description:
Video Stabilization is now considered an old problem which is almost solved but there are still some connecting problems which needs research attention. One of such issues arises due to multiple unstable videos streams coming from multiple sensors which often contain complementary information. To enhance system performance, instability should be removed in a single go rather than stabilizing each sensor individually. This paper proposes a cooperative video stabilization framework, VSAMS for multisensory aerial data based on robust boosting curves which encapsulate stability of high spatial frequency information as used by flying parakeets (budgerigars). For reducing shake and jitter and preservation of actual camera path, a multistage smoothing approach is visualized. Experiments are performed on multisensory UAV data which contains infrared and electro-optical video streams. Subjective and objective quality evaluation proves effectiveness of the proposed cooperative stabilization framework.
Description:
In this study, a population-based continuous state transition algorithm (STA) is investigated into continuous constrained optimization problems. After an analysis of the advantages and disadvantages of two well-known constraint-handling techniques, penalty function method and feasibility preference method, a two-stage strategy is proposed for constrained STA, in which, the feasibility preference method is adopted in the early stage of an iteration process whilst it is changed to the penalty function method in the later stage. Several benchmark tests are given to evaluate the performance of the proposed method, and the experimental results show that the constrained STA with a two-stage strategy outperforms other single strategy in terms of both global search ability and solution precision.
Description:
Density estimation is the ubiquitous base modelling mechanism employed for many tasks such as clustering, classification, anomaly detection and information retrieval. Commonly used density estimation methods such as kernel density estimator and k-nearest neighbour density estimator have high time and space complexities which render them inapplicable in problems with large data size and even a moderate number of dimensions. This weakness sets the fundamental limit in existing algorithms for all these tasks. We propose the first density estimation method which stretches this fundamental limit to an extent that dealing with millions of data can now be done easily and quickly. We analyze the error of the new estimation (from the true density) using a bias-variance analysis. We then perform an empirical evaluation of the proposed method by replacing existing density estimators with the new one in two current density-based algorithms, namely, DBSCAN and LOF. The results show that the new density estimation method significantly improves the runtime of DBSCAN and LOF, while maintaining or improving their task-specific performances in clustering and anomaly detection, respectively. The new method empowers these algorithms, currently limited to small data size only, to process very large databases - setting a new benchmark for what density-based algorithms can achieve.
Description:
Learning optimal Bayesian networks (BN) from data is NP-hard in general. Nevertheless, certain BN classes with additional topological constraints, such as the dynamic BN (DBN) models, widely applied in specific fields such as systems biology, can be efficiently learned in polynomial time. Such algorithms have been developed for the Bayesian-Dirichlet (BD), Minimum Description Length (MDL), and Mutual Information Test (MIT) scoring metrics. The BD-based algorithm admits a large polynomial bound, hence it is impractical for even modestly sized networks. The MDL-and MIT-based algorithms admit much smaller bounds, but require a very restrictive assumption that all variables have the same cardinality, thus significantly limiting their applicability. In this paper, we first propose an improvement to the MDL-and MIT-based algorithms, dropping the equicardinality constraint, thus significantly enhancing their generality. We also explore local Markov blanket based algorithms for constructing BN in the context of DBN, and show an interesting result: under the faithfulness assumption, the mutual information test based local Markov blanket algorithms yield the same network as learned by the global optimization MIT-based algorithm. Experimental validation on small and large scale genetic networks demonstrates the effectiveness of our proposed approaches.
Description:
Variation in viewpoints poses significant challenges to action recognition. One popular way of encoding view-invariant action representation is based on the exploitation of epipolar geometry between different views of the same action. Majority of representative work considers detection of landmark points and their tracking by assuming that motion trajectories for all landmark points on human body are available throughout the course of an action. Unfortunately, due to occlusion and noise, detection and tracking of these landmarks is not always robust. To facilitate it, some of the work assumes that such trajectories are manually marked which is a clear drawback and lacks automation introduced by computer vision. In this paper, we address this problem by proposing view invariant action matching score based on epipolar geometry between actor silhouettes, without tracking and explicit point correspondences. In addition, we explore multi-body epipolar constraint which facilitates to work on original action volumes without any pre-processing. We show that multi-body fundamental matrix captures the geometry of dynamic action scenes and helps devising an action matching score across different views without any prior segmentation of actors. Extensive experimentation on challenging view invariant action datasets shows that our approach not only removes long standing assumptions but also achieves significant improvement in recognition accuracy and retrieval.
Description:
Increase in the number of coexisting networks in Industrial, Scientific and Medical (ISM) band cause interferences and demands for intelligent interference avoidance schemes. This paper proposes a novel Diversified Adaptive Frequency Rolling (DAFR) technique for frequency hopping in Bluetooth piconets which has the tendency to mitigate both the self and static interferences and ensures sufficient frequency diversity. Simulation studies validate the prospects for the proposed scheme to be used for frequency hopping networks against already existing techniques, Adaptive Frequency Hopping (AFH) and Adaptive Frequency Rolling (AFR).
Description:
This paper presents a new LiDAR segmentation technique for automatic building detection and roof plane extraction. First, it uses a height threshold, based on the digital elevation model it divides the LiDAR point cloud into 'ground' and 'non-ground' points. Then, starting from the maximum LiDAR height, and decreasing the height at each iteration, it looks for points to form planar roof segments. At each height level, it clusters the points based on the distance and finds straight lines using the points. The nearest coplanar point to the midpoint of each line is used as a seed point and the plane is grown in a region growing fashion. Finally, a rule-based procedure is followed to remove planar segments in trees. The experimental results show that the proposed technique offers a high building detection and roof plane extraction rates while compared to other recently proposed techniques.
Description:
Doubly Fed Induction Generators (DFIGs) are nowadays extensively used in variable speed wind power plants due to their advantages that include reduced converter rating, low cost, reduced losses with an improved efficiency, easy implementation of power factor correction schemes, variable speed operation and four quadrants active and reactive power control capabilities. On the other hand, DFIG sensitivity to grid disturbances, especially for the voltage dip represents the main disadvantage of the equipment. This paper presents the application of superconducting magnetic energy storage (SMES) unit to improve the overall performance of a DFIG-based wind energy conversion systems (WECS). Two control approaches namely; hysteresis current controller (HCC) along with proportional integral (PI) controller and a fuzzy logic (FL) controller for the SMES unit are presented and compared. Various grid codes are used to examine the capability of the two proposed controllers to improve the low voltage ride through (LVRT) capability of the DFIG to maintain the wind turbine connection to the grid through fault durations.
Description:
IEEE Power and Energy Society General Meeting
Description:
Limited transmission capacity may lead to wind curtailment during periods of high availability of wind. This paper presents an improved methodology to quantify the latent scheduling capacity of a power system taking into account stochastic variation in line-thermal rating, intermittency of wind, and mitigating the risk of network congestion associated with high penetration of wind. The approach is aimed at strategic planning of power systems in the context of power systems with short to medium length lines with a priori known unit commitment decisions and uses stochastic optimization with a two stage recourse action. Results suggest that a considerable level of wind penetration is possible with dynamic line ratings, without adversely affecting the risk of network congestion.
Description:
IEEE Power and Energy Society General Meeting
Description:
In emerging 3D video coding, depth has significant importance in view synthesis, scene analysis, and 3D object reconstruction. Depth images can be characterized by sharp edges and smooth large regions. Most of the existing depth coding techniques use intra-coding mode and try to preserve edges explicitly with approximated edge modelling. However, edges can be implicitly preserved as long as the transformation is avoided. In this paper, we have demonstrated that inherent edge preserving encoding of inter-coded block residuals, uniformly quantized at pixel domain using motion data from associated texture components, is more efficient than explicitly edge preserving intra-coding techniques. Experimental results show that the proposed technique have achieved superior image quality of synthesized views against the new 3D-HEVC standard. Lossless applications of the proposed technique has achieved on average 66% and 23% bit-rate savings against 3D-HEVC with negligible quantization and perceptually unnoticeable view synthesis, respectively.
Description:
Increase in the number of coexisting networks in license free Industrial, Scientific and Medical (ISM) band causes interferences for industrial automation, e.g., shop floors of manufacturing facilities. In order to ensure the reliability for automation networks, interference avoidance schemes are required. This paper proposes a novel Predefined Hopping Pattern (PHP) technique for frequency hopping in ISM band, which mitigates self-interferences and static interferers as well. This technique generates optimized frequency hopping sequences which ensure sufficient frequency diversity and frequency offset among the coexisting Bluetooth piconets and exploits transmission experiences for a particular frequency in eliminating interference. Simulation studies have shown that PHP has better collision avoidance rate than well known adaptive frequency hopping (AFH) and adaptive frequency rolling (AFR) schemes.
Description:
The phenomenon of deadlock in robotic cells has been long ignored by most scheduling literature. A deadlock situation arises if a part cannot change its current state indefinitely since the destination machine is occupied by another part. The probability of the deadlock occurrence is likely to be large when the processing route cannot be predicted with certainty due to inspection processes. Our focus here is on a specific robotic cell with a post-process inspection system where the inspection is performed on an independent inspection machine. Avoidance and recovery policies are applied to overcome deadlocks originated from this cell. We develop these policies to prevent deadlock or alternatively resolve it during the online implementation of cycles. The former policy minimizes the storage cost, whereas the later policy minimizes the expected cycle time. An analysis of the scheduling problem that involves timings and costs is also carried out for comparing policies.
Description:
Similarity measure is an important research topic in image classification and retrieval. Given a type of image features, a good similarity measure should be able to retrieve similar images from the database while discard irrelevant images from the retrieval. Similarity measures in literature are typically distance based which measure the spatial distance between two feature vectors in high dimensional feature space. However, this type of similarity measures do not have any perceptual meaning and ignore the neighborhood influence in the similarity decision making process. In this paper, we propose a novel dissimilarity measure, which can measure both the distance and perceptual similarity of two image features in feature space. Results show the proposed similarity measure has a significant improvement over the traditional distance based similarity measure commonly used in literature.
Description:
International Conference Image and Vision Computing New Zealand
Description:
Hand vein biometric is an important modality for human authentication and liveness detection in many applications. Reliable feature extraction is vital to any biometric system. Over the past years, two major categories of vein features, namely vein structures and vein image textures, were proposed for hand dorsal vein based biometric identification. Of them, texture features seem important as it can combine skin micro-textures along with vein properties. In this study, we have performed a comparative study to identify potential texture features and feature-classifier combination that produce efficient vein biometric systems. Seven texture features (HOG, GABOR, GLCM, SSF, DWT, WPT, and LBP) and three multiclass classifiers (LDA, ESVM, and KNN) were explored towards the supervised identification of human from vein images. An experiment with 400 infrared (IR) hand images from 40 adults indicates the superior performance of the histogram of oriented gradients (HOG) and simple local statistical feature (SSF) with LDA and ESVM classifiers in terms of average accuracy (> 90%), average Fscore (> 58%) and average specificity (>93%). The decision-level fusion of the LDA and ESVM classifier with single texture features showed improved performances (by 2.2 to 13.2% of average Fscore) over individual classifier for human identification with IR hand vein images.
Description:
Proceedings - International Computer Software and Applications Conference
Description:
Hand vein biometric is an important modality for human authentication and liveness detection in many applications. Reliable feature extraction is vital to any biometric system. Over the past years, two major categories of vein features, namely vein structures and vein image textures, were proposed for hand dorsal vein based biometric identification. Of them, texture features seem important as it can combine skin micro-textures along with vein properties. In this study, we have performed a comparative study to identify potential texture features and feature-classifier combination that produce efficient vein biometric systems. Seven texture features (HOG, GABOR, GLCM, SSF, DWT, WPT, and LBP) and three multiclass classifiers (LDA, ESVM, and KNN) were explored towards the supervised identification of human from vein images. An experiment with 400 infrared (IR) hand images from 40 adults indicates the superior performance of the histogram of oriented gradients (HOG) and simple local statistical feature (SSF) with LDA and ESVM classifiers in terms of average accuracy (> 90%), average Fscore (> 58%) and average specificity (>93%). The decision-level fusion of the LDA and ESVM classifier with single texture features showed improved performances (by 2.2 to 13.2% of average Fscore) over individual classifier for human identification with IR hand vein images.
Description:
Proceedings - International Computer Software and Applications Conference
Description:
The subjective quality assessment (SQA) is an ever demanding approach due to its in-depth interactivity to the human cognition. The addition of no-reference based scheme could equip the SQA techniques to tackle further challenges. Existing widely used objective metrics-peak signal-to-noise ratio (PSNR), structural similarity index (SSIM) or the subjective estimator-mean opinion score (MOS) requires original image for quality evaluation that limits their uses for the situation having no-reference. In this work, we present a no-reference based SQA technique that could be an impressive substitute to the reference-based approaches for quality evaluation. The High Efficiency Video Coding (HEVC) reference test model (HM15.0) is first exploited to generate five different qualities of the HEVC recommended eight class sequences. To assess different aspects of coded video quality, a group of ten participants are employed and their eye-tracker (ET) recorded data demonstrate closer correlation among gaze plots for relatively better quality video contents. Therefore, we innovatively calculate the amount of approximation of smooth eye traversal (ASET) by using distance, angle, and pupil-size feature from recorded gaze trajectory data and develop a new-quality metric based on eye traversal (QMET). Experimental results show that the quality evaluation carried out by QMET is highly correlated to the HM recommended coding quality. The performance of the QMET is also compared with the PSNR and SSIM metrics to justify the effectiveness of each other.
Description:
International Conference Image and Vision Computing New Zealand
Description:
The subjective quality assessment (SQA) is an ever demanding approach due to its in-depth interactivity to the human cognition. The addition of no-reference based scheme could equip the SQA techniques to tackle further challenges. Existing widely used objective metrics-peak signal-to-noise ratio (PSNR), structural similarity index (SSIM) or the subjective estimator-mean opinion score (MOS) requires original image for quality evaluation that limits their uses for the situation having no-reference. In this work, we present a no-reference based SQA technique that could be an impressive substitute to the reference-based approaches for quality evaluation. The High Efficiency Video Coding (HEVC) reference test model (HM15.0) is first exploited to generate five different qualities of the HEVC recommended eight class sequences. To assess different aspects of coded video quality, a group of ten participants are employed and their eye-tracker (ET) recorded data demonstrate closer correlation among gaze plots for relatively better quality video contents. Therefore, we innovatively calculate the amount of approximation of smooth eye traversal (ASET) by using distance, angle, and pupil-size feature from recorded gaze trajectory data and develop a new-quality metric based on eye traversal (QMET). Experimental results show that the quality evaluation carried out by QMET is highly correlated to the HM recommended coding quality. The performance of the QMET is also compared with the PSNR and SSIM metrics to justify the effectiveness of each other.
Description:
International Conference Image and Vision Computing New Zealand