Computer Scince

Due to the limitation of computing complexity, it is difficult to apply the H.264 deblocking filter to low-end terminals. Although some technologies to optimize it have been proposed, the complexity is still high for real time implementation. Considering that deblocking filter is applied to all of the vertical and horizontal edges of 4x4 blocks but only some of them actually need to be filtered , in this paper, based on previous encoding or decoding information, we propose a novel deblocking filter algorithm to reduce complexity of exist algorithms by skipping quantities of needless edges. The experimental results show that our proposed algorithm has better performance in subjective video quality and time efficiency compared to JM 10.2 algorithm.

Reconstruction of hyperspectral imagery from spectral random projections is considered. Specifically, multiple predictions drawn for a pixel vector of interest are made from spatially neighboring pixel vectors within an initial non-predicted reconstruction. A two-phase hypothesis-generation procedure based on partitioning and merging of spectral bands according to the correlation coefficients between bands is proposed to fine-tune the hypotheses. The resulting prediction is used to generate a residual in the projection domain. This residual being typically more compressible than the original pixel vector leads to improved reconstruction quality. To appropriately weight the hypothesis predictions, a distance-weighted Tikhonov regularization to an ill-posed least-squares optimization is proposed. Experimental results demonstrate that the proposed reconstruction significantly outperforms alternative strategies not employing multihypothesis prediction.

Compressed-sensing reconstruction of still images and video sequences driven by multihypothesis predictions is considered. Specifically, for still images, multiple predictions drawn for an image block are made from spatially surrounding blocks within an initial non-predicted reconstruction. For video, multihypothesis predictions of the current frame are generated from one or more previously reconstructed reference frames. In each case, the predictions are used to generate a residual in the domain of the compressed-sensing random projections. This residual being typically more compressible than the original signal leads to improved reconstruction quality. To appropriately weight the hypothesis predictions, a Tikhonov regularization to an ill-posed least-squares optimization is proposed. Experimental results demonstrate that the proposed reconstructions outperform alternative strategies not employing multihypothesis predictions. 978-1-4673-0323-1/11/$26.00

Single-image super-resolution driven by multihypothesis prediction is considered. The proposed strategy exploits self-similarities existing between image patches within a single image. Specifically, each patch of a low-resolution image is represented as a linear combination of spatially surrounding hypothesis patches. The coefficients of this representation are calculated using Tikhonov regularization and then used to generate a high-resolution image. Experimental results reveal that the proposed algorithm offers significantly higher-quality super-resolution than bicubic interpolation without the cost of training on an extensive training set of imagery as is typical of competing single-image techniques.

This paper emphasizes particularly on introduction of the application of non-Redundant Rules Algorithm on Data Analyses of Forest Inventory. By establishing the data mining model, MVNR Algorithm is applied to analyzing the relation of species, origin, age, chest, circumference, height and canopy density of trees. The results provide the best valuable information for forestation programming management. The promising prospect of the application of this method in the investigation and analysis of data for the forestry resources has been proved.

The new generation video coding standard H.264/AVC has been developed to achieve higher coding efficiency than previous video coding standards. Its variable block size motion compensation requires huge computation, which is a big problem for real time application. To reduce the complexity, we propose a fast inter mode decision algorithm with early mode prediction rules and early block size classification by using RD cost of the mode. Only parts of the inter prediction modes are chosen for motion estimation and rate distortion optimization. An adaptive threshold makes the proposed algorithm works smooth for different video sequences with a wide variety of activity. The experimental results indicate that the proposed algorithm can achieve 53.9% time saving in inter mode decision on average without any noticeable performance degradation, and the increased of bit rate is less than 1%.

This paper presents a human action recognition method by using depth motion maps. Each depth frame in a depth video sequence is projected onto three orthogonal Cartesian planes. Under each projection view, the absolute difference between two consecutive projected maps is accumulated through an entire depth video sequence forming a depth motion map. An 2 l  regularized collaborative representation classifier with a distance-weighted Tikhonov matrix is then employed for action recognition. The developed method is shown to be computationally efficient allowing it to run in real-time. The recognition results applied to the Microsoft Research Action3D dataset indicate superior performance of our method over the existing methods.

The Wireless Sensor Network technology has been used widely; however the limited energy resource is one of the bottlenecks for its application. To enhance the robustness and accuracy of the information which is obtained by the entire network, certain redundancy exists in the data collected from the sensor node and thus data fusion processing is needed to reduce the redundant information. Firstly the function and categories of data fusion were introduced in this paper, and also the relation between data fusion and protocol layers of the sensor network were described. Meanwhile various data fusions which combine the route patterns in the network, especially dynamic cluster organization algorithm were emphasized on. On the basis of analyzing the principle of the LEACH protocol, a recent typical data fusion algorithm, the simulation and contrastive analysis of the LEACH protocol and its ameliorated algorithm, the LEACH-C protocol, were implemented by using the NS2 network simulation software. The simulation results suggest that the LEACH-C protocol can effectively solve the problems such as the uncertainty of the number of the cluster header in the LEACH algorithm, lack of consideration of the energy state of the nodes in the construction stage of dynamic cluster and so on.

Spectral-spatial preprocessing using multihypothesis prediction is proposed for improving accuracy of hyperspectral image classification. Specifically, multiple spatially collocated pixel vectors are used as a hypothesis set from which a prediction for each pixel vector of interest is generated. Additionally, a spectral-bandpartitioning strategy based on inter-band correlation coefficients is proposed to improve the representational power of the hypothesis set. To calculate an optimal linear combination of the hypothesis predictions, a distance-weighted Tikhonov regularization to an illposed least-squares optimization is used. The resulting predictions effectively integrate spectral and spatial information and thus are used during classification in lieu of the original pixel vectors. This processed hyperspectral image dataset has less intraclass variability and more spatial regularity as compared to the original dataset. Classification results for two hyperspectral image datasets demonstrate that the proposed method can enhance the classification accuracy of both maximum-likelihood and support vector classifiers, especially under small sample size constraints and noise corruption.

To  address  the  PID  parameter  tuning  problem,  inspired  by  the  swarm  intelligence
optimization algorithm, a tuning method of PID parameters based on particle swarm algorithm is
proposed. To find an optimal set of PID control parameters in the target space, three parameters of the
PID  are  as  particles  based  on  the  fitness  function.  By  designing  optimization  programs,
air-conditioning temperature control system experiments were carried out as an example. The results
show that the optimization can improve the dynamic control performance of the system.

Applying  EKF(extended  Kalman  filter)  based  integrated  navigation  to  correct  the  measurement  error  of  MEMS
（Micro-electromechanical  Systems ）suffers  great  computational  load,  slow  convergence  and  implementation
infeasibility. Based  on  steepest  decent  algorithm and  complementary filter  theory,  this  paper  presents  an  attitude
error correction method using MARG (Magnetic, Angular Rate, and Gravity) sensor array. This method uses the
Jacobian  matrix  to  substitute  the  recurrence  formula  of  EKF,  which  significantly  reduces  the  computational
complexity.  It  only  requires  to  tune  two  filter  gains  which  are  determined  by  the  gyroscope  measurement  error
leading  to  an  ease  of  implementation.  The  experimental  results  of  MARG/EKF  simulation  and  MARG/turntable
indicate  that  our  proposed  method  has  lower  computational  complexity  and  faster  convergence  than  the  EKF
method.  In  addition,  our  proposed  method  can  achieve  high  accuracy  when  MARG  sensors  are  operated  at  low
sampling  rates,  which  is  very  desirable  for  applications  constrained  by  low  power  consumption  or  limited
processing resources.

PID control is widely used in air conditioning
system. The algorithm is simple but has a long response
time, and the adjustment is not flexible. In this paper, we
propose a method of optimizing PID controller parameters
based on adaptive genetic algorithm. Specifically, the
fitness function with penalty terms is added to the classical
genetic algorithm and adaptive crossover and mutation
factor genes to make the PID parameters converge to
global optimum quickly. In addition, the method can
prevent overshoot and improve stability effectively.
Simulation results indicate that our method achieved a
83% reduction in regulation time of response of an airconditioning unit compared with the original PID system.
The method can be used for online tuning of PID
parameters as well.

Extreme learning machine (ELM) is a single-layer feedforward neural network based classifier that has attracted significant attention in computer vision and pattern recognition due to its fast learning speed and strong generalization. In this paper, we propose to integrate spectral-spatial information for hyperspectral image classification and exploit the benefits of using spatial features for the kernel based ELM (KELM) classifier. Specifically, Gabor filtering and multihypothesis (MH) prediction preprocessing are two approaches employed for spatial feature extraction. Gabor features have currently been successfully applied for hyperspectral image analysis due to the ability to represent useful spatial information. MH prediction preprocessing makes use of the spatial piecewise-continuous nature of hyperspectral imagery to integrate spectral and spatial information. The proposed Gabor-filtering-based KELM classifier and MH-prediction-based KELM classifier have been validated on two real hyperspectral datasets. Classification results demonstrate that the proposed methods outperform the conventional pixel-wise classifiers as well as Gabor-filtering-based support vector machine (SVM) and MH-prediction-based SVM in challenging small training sample size conditions. OPEN ACCESS Remote Sens. 2014, 6 5796

This paper presents a new dynamic classifier selection approach for hyperspectral image classification, in which both spatial and spectral information are used to determine a pixel's label once the remaining classified pixels' neighborhood meets the threshold. For volumetric texture feature extraction, a volumetric gray level co-occurrence matrix is used; for spectral feature extraction, a minimum estimated abundance covariance-based band selection is used. Two hyperspectral remote sensing datasets, HYDICE Washington DC Mall and AVIRIS Indian Pines, are employed to evaluate the performance of the developed method. The classification accuracies of the two datasets are improved by 1.13% and 4.47%, respectively, compared with the traditional algorithms using spectral information. The experimental results demonstrate that the integration of spectral information with volumetric textural features can improve the classification performance for hyperspectral images.

This paper presents a home-based Senior Fitness Test (SFT) measurement system by using an inertial sensor and a depth camera in a collaborative way. The depth camera is used to monitor the correct pose of a subject for a fitness test and any deviation from the correct pose while the inertial sensor is used to measure the number of a fitness test action performed by the subject within the time duration specified by the fitness protocol. The results indicate that this collaborative approach leads to high success rates in providing the SFT measurements under realistic conditions.

This paper presents a medication adherence monitoring system for pill bottles based on a wearable inertial sensor. Signal templates corresponding to the two actions of twist-cap and hand-to-mouth are created using a camera-assisted training phase. The act of pill intake is then identified by performing a moving window dynamic time warping in real-time between signal templates and the signals acquired by the wearable inertial sensor. The outcomes of the experimentations carried out indicate that the developed medical adherence monitoring system identifies the act of pill intake with a high degree of accuracy.

In view of the characteristics of central air conditioning control system (with great inertia and multiple disturbances) and the prematurity of conventional genetic algorithm, this  paper integrated simulated annealing algorithm into conventional genetic algorithm to realise the
proportional, integral, differential (PID) parameter tuning. Local optimal solution was effectively rejected by stretching fitness function and introducing access criterion with controlled probability. MATLAB simulation results showed that the system response lagged only 1000s,
and the efficiency was improved by 80% by using this method. In addition, the fitness value was smaller than that of conventional genetic algorithm.

This paper presents a multi-Hidden Markov Model (HMM) classification approach for hand gesture recognition by utilizing two differing modality and low-cost sensors. The sensors consist of a Kinect depth camera and a wearable inertial sensor. It is shown that the multi-HMM classification based on nine signals that are simultaneously captured by these two sensors leads to a more robust recognition compared to the situation when only a single HMM classification is used to generate the likelihood probabilities of hand gestures. This approach is applied to the hand gestures of the $1Unistroke Recognizer application and the results obtained indicate a 7% improvement in the overall classification rate over a single HMM classification under realistic conditions.