Aiming at the correlation problem between sleep stages and sleep apnea hypopnea， a sleep physiological time series classification method based on adaptive multi-task learning was proposed. Single-channel electroencephalogram and electrocardiogram were used for sleep staging and Sleep Apnea Hypopnea Syndrome （SAHS） detection. A two-stream time dependence learning module was utilized to extract shared features under joint supervision of the two tasks. The correlation between sleep stages and sleep apnea hypopnea was modeled by the adaptive inter-task correlation learning module with channel attention mechanism. The experimental results on two public datasets indicate that the proposed method can complete sleep staging and SAHS detection simultaneously. On UCD dataset， the accuracy， MF1（Macro F1-score）， and Area Under the receiver characteristic Curve （AUC） for sleep staging of the proposed method were 1.21 percentage points， 1.22 percentage points， and 0.008 3 higher than those of TinySleepNet； its MF2 （Macro F2-score）， AUC， and recall of SAHS detection were 11.08 percentage points， 0.053 7， and 15.75 percentage points higher than those of the 6-layer CNN model， which meant more disease segments could be detected. The proposed method could be applied to home sleep monitoring or mobile medical to achieve efficient and convenient sleep quality assessment， assisting doctors in preliminary diagnosis of SAHS.

With the development of Artificial Intelligence （AI） technology， the Deep Neural Network （DNN） models have been applied to various mobile and edge devices widely. However， the model deployment becomes challenging and the popularization and application of the models are limited due to the facts that the computing power of edge devices is low， the memory capacity of edge devices is small， and the realization of model acceleration requires in-depth knowledge of edge device hardware. Therefore， a DNN acceleration and deployment method based on Tensor Virtual Machine （TVM） was presented to accelerate the Convolutional Neural Network （CNN） model on Field-Programmable Gate Array （FPGA）， and the feasibility of this method was verified in the application scenarios of distracted driving classification. Specifically， in the proposed method， the computational graph optimization method was utilized to reduce the memory access and computational overhead of the model， the model quantization method was used to reduce the model size， and the computational graph packing method was adopted to offload the convolution calculation to the FPGA in order to speed up the model inference. Compared with MPU （MicroProcessor Unit）， the proposed method can reduce the inference time of ResNet50 and ResNet18 on MPU+FPGA by 88.63% and 77.53% respectively. On AUC （American University in Cairo） dataset， compared to MPU， the top1 inference accuracies of the two models on MPU+FPGA are only reduced by 0.26 and 0.16 percentage points respectively. It can be seen that the proposed method can reduce the deployment difficulty of different models on FPGA.

Accurate prediction of taxi demands between urban regions can provide decision support information for taxi guidance and scheduling as well as passenger travel recommendation， so as to optimize the relation between taxi supply and demand. However， most of the existing models only focus on modeling and predicting the taxi demands within a region， do not consider enough the spatial-temporal correlation between regions， and pay less attention to the more fine-grained demand prediction between regions. To solve the above problems， a prediction model for taxi demands between urban regions — Origin-Destination fusion with Spatial-Temporal Network （ODSTN） model was proposed. In this model， complex spatial-temporal correlations between regions was captured from spatial dimensions of the regions and region pairs respectively and three temporal dimensions of recent， daily and weekly periods by using graph convolution and attention mechanism， and a new path perception fusion mechanism was designed to combine the multi-angle features and finally realize the taxi demand prediction between urban regions. Experiments were carried out on two real taxi order datasets in Chengdu and Manhattan. The results show that the Mean Absolute Error （MAE）， Root Mean Square Error （RMSE） and Mean Absolute Percentage Error （MAPE） of ODSTN model are 0.897 1， 3.527 4， 50.655 6% and 0.589 6， 1.163 8， 61.079 4%， respectively， indicating that ODSTN model has high accuracy in taxi demand prediction tasks.

The complex structure and diverse imformation of stock markets make stock movement prediction extremely challenging. However， most of the existing studies treat each stock as an individual or use graph structures to model complex higher-order relationships in stock markets， without considering the hierarchy and dynamics among stocks， industries and markets. Aiming at the above problems， a Dynamic Macro Memory Network （DMMN） was proposed， and price movement prediction was performed for multiple stocks simultaneously based on DMMN. In this method， the market macro-environmental information was modeled by the hierarchies of “stock-industry-market”， and long-term dependences of this information on time series were captured. Then， the market macro-environmental information was integrated with stock micro-characteristic information dynamically to enhance the ability of each stock to perceive the overall state of the market and capture the interdependences among stocks， industries， and markets indirectly. Experimental results on the collected CSI300 dataset show that compared with stock prediction methods based on Attentive Long Short-Term Memory （ALSTM） network， GCN-LSTM （Graph Convolutional Network with Long Short-Term Memory）， Convolutional Neural Network （CNN） and other models， the DMMN-based method achieves better results in F1-score and Sharpe ratio， which are improved by 4.87% and 31.90% respectively compared with ALSTM， the best model among all comparison methods. This indicates that DMMN has better prediction performance and better practicability.

Traffic flow forecasting is an important research topic for the intelligent transportation system， however， this research is very challenging because of the complex local spatio-temporal relationships among traffic objects such as stations and sensors. Although some previous studies have made great progress by transforming the traffic flow forecasting problem into a spatio-temporal graph forecasting problem， in which the direct correlations across spatio-temporal dimensions among traffic objects are ignored. At present， there is still lack of a comprehensive modeling approach for the local spatio-temporal relationships. A novel spatio-temporal hypergraph modeling scheme was first proposed to address this problem by constructing a kind of spatio-temporal hyper-relationships to comprehensively model the complex local spatio-temporal relationships. Then， a Spatio-Temporal Hyper-Relationship Graph Convolutional Network （STHGCN） forecasting model was proposed to capture these relationships for traffic flow forecasting. Extensive comparative experiments were conducted on four public traffic datasets. Experimental results show that compared with the spatio-temporal forecasting models such as Attention based Spatial-Temporal Graph Convolutional Network （ASTGCN） and Spatial-Temporal Synchronous Graph Convolutional Network （STSGCN）， STHGCN achieves better results in Root Mean Square Error （RMSE）， Mean Absolute Error （MAE） and Mean Absolute Percentage Error （MAPE）； and the comparison of the running time of different models also shows that STHGCN has higher inference speed.

In order to overcome the problems of low convergence precision and easily relapsing into local optimum in Fruit fly Optimization Algorithm (FOA), by introducing the Levy flight strategy into the FOA, an improved FOA called double subgroups FOA with the characteristics of Levy flight (LFOA) was proposed. Firstly, the fruit fly group was dynamically divided into two subgroups (advanced subgroup and drawback subgroup) whose centers separately were the best individual and the worst individual in contemporary group according to its own evolutionary level. Secondly, a global search was made for drawback subgroup with the guidance of the best individual, and a finely local search was made for advanced subgroup by doing Levy flight around the best individual, so that not only both the global and local search ability balanced, but also the occasionally long distance jump of Levy flight could be used to help the fruit fly jump out of local optimum. Finally, two subgroups exchange information by updating the overall optimum and recombining the subgroups. The experiment results of 6 typical functions show that the new method has the advantages of better global searching ability, faster convergence and more precise convergence.

Concerning the contradiction between edge-preserving and noise-suppressing in the process of image denoising, a patch similarity anisotropic diffusion algorithm based on variable exponent for image denoising was proposed. The algorithm combined adaptive Perona-Malik (PM) model based on variable exponent for image denoising and the idea of patch similarity, constructed a new edge indicator and a new diffusion coefficient function. The traditional anisotropic diffusion algorithms for image denoising based on the intensity similarity of each single pixel (or gradient information) to detect edge cannot effectively preserve weak edges and details such as texture. However, the proposed algorithm can preserve more detail information while removing the noise, since the algorithm utilizes the intensity similarity of neighbor pixels. The simulation results show that, compared with the traditional image denoising algorithms based on Partial Differential Equation (PDE), the proposed algorithm improves Signal-to-Noise ratio (SNR) and Peak-Signal-to-Noise Ratio (PSNR) to 16.602480dB and 31.284672dB respectively, and enhances anti-noise capability. At the same time, the filtered image preserves more detail features such as weak edges and textures and has good visual effects. Therefore, the algorithm achieves a good balance between noise reduction and edge maintenance.

A task scheduling strategy based on the dynamically adaptive ant colony algorithm was proposed for the first time to solve the drawbacks like slow convergence and easily falling into local optimal that have long existed in the ant colony algorithm. Chaos disruption was introduced when selecting the resource node, the pheromone evaporation factors were adjusted adaptively based on nodes pheromone and the pheromone were updated dynamically according to the solutions performance. When the number of tasks was greater than 150, compared with the dynamically adaptive ant colony algorithm and ant colony algorithm, time efficiency could be maximally improved up to 319% and resource load was 0.51.The simulation results prove that the proposed algorithm is suitable for improving convergence rate and the global searching ability.

For the purpose of correcting the distortion of the large-field-of-view lens used on aerial cameras, this paper proposed a method using the Matlab calib_toolbox. By calibrating the captured images of a range of angles-of-view and distances, the internal parameters and distortion coefficients of the camera were obtained, and the correct mathematical model of distortion correction was built. The proposed method made an improvement on the Bouguet method, and extended its applications to distortion correction of color images for airborne cameras via subsequent programming. Furthermore, a new and efficient backstepping reconstruction pattern matching method for image-distortion-rate analysis was proposed, which quantified the level of distortion. The simulation results show that the proposed method reduces the distortion rate of color images averagely by about 10%. As all the experimental results indicate, the proposed method is simple and efficient, and it is convenient to be readily transplanted to hardware platform for real-time image distortion correction.

In response to the issue that GPS is unable to carry out Location-Based Service (LBS) in indoor environment, a LBS indoor channel allocation model, credibility evaluation and control method was presented in this paper, which integrated GPS, Wi-Fi, ZigBee and Bluetooth technologies. It solved the problem arising from combination channel allocation, including the evaluation of the traffic load, the available Radio Frequency (RF), and non-overlapping RF channels number of each node. Each Access Point(AP)'s signal strength built the prediction model with reference point. The optimization algorithm was designed to determine and select the credibility of combination channel based on the energy evaluation. It adaptively selected neighbors with highest comprehensive effects to participate in iterative optimization. The simulation result indicates this method can effectively inhibit the proliferation of communication interference error in the network, reduce the positioning complexity, and improve the positioning accuracy in addition to improving scalability and robustness of the entire network.

In a Virtual Machine (VM) computing environment, it is difficult to monitor and allocate the VM's memory in real-time. To overcome these shortcomings, a real-time method of monitoring and adjusting memory for Xen virtual machine called Xen Memory Monitor and Control (XMMC) was proposed and implemented. This method used hypercall of Xen, which could not only real-time monitor the VM's memory usage, but also dynamically real-time allocated the VM's memory by demand. The experimental results show that XMMC only causes a very small performance loss, less than 5%, to VM's applications. It can real-time monitor and adjust on demand VM's memory resource occupations, which provides convenience for the management of multiple virtual machines.

Because of its unique advantage of information propagation, the online social network has been a popular social communication platform. In view of the characteristics of the form of information propagation and the dynamics theory of infectious diseases, this paper put forward the model of information propagation through online social network. The model considered the influence of different users' behaviors on the transmission mechanism, set up the evolution equations of different user nodes, simulated the process of information propagation, and analyzed the behavior characteristics of the different types of users and main factors that influenced the information propagation. The experimental results show that different types of users have special behavior rules in the process of information propagation, i.e., information cannot be transported endlessly, and be reached at a stationary state, and the larger the spread coefficient or immune coefficient is, the faster it reached the stationary state.

The limited battery life is a big obstacle for the further growth of mobile devices, so a new dynamic application partitioning algorithm was proposed to minimize power consumption of mobile devices by offloading its computation to a remote resource-rich server. An Object Relation Graph (ORG) for such an application was set up, and then it was transformed into a network. By using network flow theory, the optimization problem of power consumption was transformed into the optimal bipartition problem of a flow network which can be partitioned by the max-flow min-cut algorithm. The simulation results show that the proposed algorithm can greatly save more energy than the existing algorithms, and better adapt to environment changes.

In view of the characteristics of Ka-band mobile satellite channel, a model was built up with taking full consideration of the decline influenced by weather and surrounding environment in this paper. Then, the satellite channel with Bipolarity Phase Shift Keying (BPSK) modulation was simulated, and the probability model of baseband channel bit error distribution was established. Meanwhile, the probability of error occurrences was fitted by using the least-squares method. According to the results, the occurrence number of channel errors obeys Poisson distribution. Moreover, during the digital baseband simulation of Ka-band mobile satellite channels, it is found that the system error with BPSK modulation consists of the burst error and the random error that obeys the Poisson distribution.

Image-based fire detection can effectively solve the problems of large space fire detection contactlessly and rapidly. It is a new research direction in fire detection. Its essential issue is the classification of flames and disruptors. The ordinary detection methods are to extract one or a few characteristics of the flame in the image as a basis for identification. The disadvantages are to need a large number of experiential thresholds and the lower recognition rate by the inappropriate feature selection. Considering the entire characteristics of fire flame, a flame detection method based on Independent Component Analysis (ICA) and Support Vector Machine (SVM) was proposed. Firstly, a series of frames were pre-processed in RGB space. And suspected target areas were extracted depending on the flickering feature and fuzzy clustering analysis. Then the flame image features were described with ICA. Finally, SVM model was used in order to achieve flame recognition. The experimental result shows that the proposed method improves the accuracy and speed of image fire detection in a variety of fire detection environments.

With the development of image retrieval system, the effective organizing and managing of image database has been a key to users retrieving. This paper firstly made use of colony clustering algorithm in emotional clustering which based on image feature, and the original ant clustering algorithm was improved. This algorithm may identify which is the first ants by calculating the continental distance, simulate ants' actions of pickup or discarding, and according to the feature of domain color, image emotional cluster has been implemented. The result of experiment has indicated that the algorithm can get better clustering effect, and improve the retrieval efficiency.

Classic decision tree algorithm is unfit to cope with uncertain data pervaded at both the construction and classification phase. In order to overcome these limitations， D-S decision tree classification algorithm was proposed. This algorithm extended the decision tree technique to an uncertain environment. To avoid the combinatorial explosion that would result from tree construction phase, uncertainty measure operator and aggregation combination operator were introduced. This D-S decision tree is a new classification method applied to uncertain data and shows good performance and can efficiently avoid combinatorial explosion.

In order to obtain the SAR images which include the typical target of interest, a new method of SAR images screening based on bit-plane characteristics was proposed according to the imaging characteristic of target. Based on the suitable gray pretreatment to the images, the target’s prior knowledge was analyzed, the significant bit-plane image was paid attention by the measurement of bit-plane complexity, run length and frequency spectrum. And then SAR images were screened combined with the gray histogram features. Around the airport SAR images, experiment shows that the method can screen the images rapidly. Besides, the airport target is extracted successfully, which can satisfy the requirements.

In view of the particularity of binary image, a digital watermarking algorithm was proposed for binary image. Watermark was put into an original image by using Arnold transformation and flipping the pixels which met the visual constraints. Experimental results show that the algorithm not only improves the unobtrusiveness and embedding capacity of watermark, but also realizes the blind watermarking.

Context-Based Adaptive Variable-Length Coding (CAVLC) algorithm was adopted as an entropy coding method in baseline and extended profile of H.264/AVC standard, but the detailed syntax on which was not explicitly stipulated. A profound analysis on the CAVLC coding algorithm of H.264 standard was performed based on the principle of CALVC decoding method. A high-speed and low power-consumption CAVLC coder for H.264/AVC standard was presented according to the former analysis, in which multi-clock domain processing and parallel processing techniques were adopted to improve the performance of the system, and arithmetic were used to replace some static code table to reduce memory consumption. The detailed design and FPGA realization method on each sub-blocks are also concerned. Finally, FPGA verification and realization indicates that the maximum coding system clock can up to 107.97MHz, and the coding delay is less than 36 clock cycles, which can adequately meet the needs of some high-definition and real-time applications.

Having done the comparative analysis of the several existing vortex detection algorithms, pattern matching algorithm based on Clifford convolution was studied with emphasis. Considering the irregularity of the data sets obtained by the actual computation, an improved method was proposed for the pattern matching algorithm based on Clifford convolution. Irregular data sets were divided by the hybrid grid and the pattern was scaled according to the cells surrounding grid point. In computation process, the 1-neighborhood of pattern was sampled. The result show that this algorithm can more accurately detect and visualize the vortices of flow field preconditioned with similar algorithm efficiency.