Multimedia Technology and Enhanced Learning

To identify tea categories more automatically and efficiently, we proposed an improved tea identification system based on Histogram Equalization (HE), Gray-Level Co-Occurrence Matrix (GLCM) and Support Vector Machine (SVM) algorithm. In our previous project, 25 images per class might be not enough to classify, and a small size of dataset will cause overfitting. Therefore, we collected 10 kinds of typical processed Chinese tea, photographed 300 images each category by Canon EOS 80D camera, and regarded them as a first-hand dataset. The dataset was randomly divided into training set and testing set, which both contain 1500 images. And we applied data augmentation methods to augment the training set to a 9000-image training set. All the images were resized to 256 * 256 pixels as the input of feature extraction process. We enhanced the image features through Histogram Equalization (HE) and extracted features from each image which were trained through Gray-Level Co-Occurrence Matrix (GLCM). The results show that the average accuracy reached 94.64%. The proposed method is effective for tea identification process.

The automobile industry is one of the most promising industries in China. The increase of the number of vehicles is accompanied by the increase of traffic accidents. Therefore, the application of composite pipe materials has a broad prospect and become the basis of the automobile industry. In this paper, the durability of the composite material is tested, and it is used in the manufacture of stainless steel pipe fittings of automobile, which can produce low-cost, high-performance and ideal automobile parts. The application of the composite material in automobile field can effectively improve the production efficiency and use efficiency.

In order to improve the database management and scheduling ability of smartphone terminal, it is necessary to optimize the information extraction of smartphone terminal database in lock screen mode. through the dynamic mining of smartphone terminal database data, the optimal processing of smart phone terminal database data information is realized, and this paper proposes an algorithm to extract the resource information of the smartphone terminal database under the lock screen mode based on the Internet. Using distributed wireless sensor to form the Internet model of smart phone database collection, and the optimal deployment design of data acquisition Internet of things node is carried out. The spectrum analysis method is used to detect the abnormal resource information of smart phone terminal database in lock screen mode, and the detection results are fuzzy clustering to realize the extraction of smart phone terminal database resource information in locked screen mode under the environment of Internet of things. The simulation results show that the algorithm has high accuracy, good recall, strong anti-interference ability and good adaptive ability to collect the resource information of smart phone terminal database in lock screen mode.

In order to improve the quality detection ability of thermal insulation building material, a phase change energy storage thermal insulation building material quality detection method based on neural network is put forward. The double broken line model is used to detect and evaluate the quality of phase change energy storage thermal insulation building material. Combined with the material stress characteristics and dissipation characteristics, the seismic evaluation of phase change energy storage thermal insulation building material is carried out, and the self-recovery energy dissipation support model of phase change energy storage thermal insulation building material is established. Under the constraint of quality parameter constraint model and neural network control model, the quality of phase change energy storage thermal insulation building materials is evaluated quantitatively by using network adaptive control method, and the stress distribution characteristic parameters of phase change energy storage thermal insulation building materials are calculated respectively to realize the quality detection of phase change energy storage thermal insulation building materials. The simulation results show that the accuracy of phase change energy storage and thermal insulation building material quality detection is high, the accuracy of parameter evaluation is good, and the quality detection ability is very good.

To improve that automatic monitoring capability of the vehicle’s pollutant discharge, a design method of the auto-monitor system for the emission of vehicle pollutant is proposed based on the fusion algorithm. The system design includes a large data analysis model design and system software for automatic monitoring of vehicle pollutant discharge. In that method, a quantitative statistical analysis method is adopt to carry out cloud computing fusion processing of automatic monitoring of pollutant discharge of a motor vehicle, and the invention relates to an information fusion and recombination model for automatically monitoring large data of pollutant discharge of a motor vehicle, the method of automatic monitoring of vehicle pollutant discharge based on the fusion algorithm is improved by the effective feature of the automatic monitoring of the pollutant discharge of the motor vehicle. In the development of B/S and embedded PCI bus, the software development and design of the automatic monitoring system for pollutant discharge of motor vehicle are carried out, and the hardware structure and software development of the automatic monitoring system for pollutant discharge of motor vehicle are carried out. The results of the test show that the automatic monitoring system for motor vehicle pollutant discharge is good in stability and strong in the sampling and statistical analysis of the physical information.

Multi-heterogeneous sensor information fusion under traditional technology conditions was slow, inaccurate, incomplete, and inconsistent, which led to errors in data analysis and affect evaluation results. To this end, IoT technology was used to study the multi-source heterogeneous sensor information fusion method. Four methods of data acquisition, data abstraction and access, feature fusion algorithm design of high attribute dimension data, and feature level information fusion method were used to creatively change the traditional operation method. The experiment proved that the IoT data information presented new characteristics under the universal characteristics of the Internet of Things, and used the high-level knowledge evolution mechanism of the information resource development chain to study the state evolution of the Internet of Things information in its life cycle. The mechanism was to customize the guiding strategy for the integration of high-quality information in the Internet of Things.

For the sake of improving the optimal management and dispatching ability of Guangdong-Hong Kong-Macao Greater Bay Area’s economy, it is essential to optimize and predict the growth trend of the Greater Bay Area’s economy, put forward the optimization prediction method of the Greater Bay Area economic growth trend based on 3500 mining, and construct the economic growth model of statistical sequence distribution. Big data mining method is chosen to model the big data statistical information of the area’s economic growth, extract the characteristic quantity of the association rules of the big data economic growth trend, use the fuzzy fusion clustering method to carry on the automatic clustering processing to the economic growth trend, and establish the optimal iterative model of the prediction of the economic growth trend. Combined with adaptive optimization algorithm, the Greater Bay Area’s economic growth trend is optimized and predicted. The simulation outputs show that the method has good adaptability to predict economic growth trend of the area we talked about, and has high accuracy in predicting growth trend, which improves the adaptive scheduling and management ability of the economy in the bay area.

China realize the sustainable development of socialism, and promote the integration and development of Guangdong, Hong Kong, Macao and the Gulf region. This paper analyzes the development path of the urban agglomeration in Da Wan District of Guangdong, Hong Kong and Macao under the backdrop of big data. A statistical sequence distribution model of the GDP index of the city group development in the Big Gulf Region of These city is constructed, the big data statistical information model of the GDP index of the city group development in the Big Gulf Region of These city is built by using a big data mining method, the association rule characteristic quantity of the GDP index of the city group development in the Big Gulf Region of These city is extracted, the big data of the GDP index of the city group development in the Big Gulf Region of These city under the big An optimization iteration model for the prediction of the GDP index for the development of the large bay area urban agglomeration in These city is established. Under the backdrop of big data, the development path analysis and adaptive adjustment of the large bay area urban agglomeration in These city are carried out to realize the analysis and optimization of the development path of the large bay area urban agglomeration The simulation results show that the prediction accuracy of the GDP index of These city and the gulf city assembly development is high, and the adaptability and convergence of the GDP index prediction of These city and the gulf city assembly development are improved.

In order to improve the fault detection capability of large and medium-sized pump stations, the abnormal feature diagnosis of the fault data is required, and the intelligent diagnosis algorithm of the fault data of the large and medium-sized pump station under the information-based evaluation system is put forward. The fault data sensing information acquisition node distribution model of the large and medium-sized pump station is constructed, the multi-sensor fusion sampling method is adopted to sample the fault data of the large and medium-sized pump station, and the statistical feature quantity of the fault data of the large and medium-sized pump station is extracted. The fault data set of large and medium-sized pump station is used to detect and optimize the abnormal working condition of the fault data set of the large and medium-sized pump station, and the fault diagnosis of the large and medium-sized pump station is realized according to the detection result. The simulation results show that the accuracy of the fault data set of large and medium pump station is high, and the real-time and self-adaptability of the fault detection are better.

Power transformer insulation fault location is the key to improve the stability of power transformer. A Bayesian network based on power transformer insulation fault on-line monitoring method is proposed. The Bayesian network characteristic decomposition model is used to detect the insulation fault of power transformer, the high-resolution spectrum characteristic quantity of insulation fault of power transformer is extracted, the load balance analysis is carried out according to the output voltage and load difference of power transformer, the Bayesian network detection model of insulation fault of power transformer is constructed. Combined with PCI integrated information processor and relay transmission node network topology model, the on-line monitoring system design of power transformer insulation failure is realized. The simulation results show that the fault location of power transformer insulation is accurate and the visual resolution of fault is strong.

Aiming at the problem of poor robustness of traditional user behavior pattern mining analysis method, a cloud computing-based intelligent home user behavior pattern mining analysis method is designed. Intelligent household IoT from using cloud computing method and data mining the user behavior patterns, establish a two-layer neural network level of data is divided into 2 kinds, the user behavior mode by setting the input weight vector calculation after classifying data correlation between user behavior model, using Apriori algorithm, input minimum support and minimum confidence, on the basis of analyzing the correlation between data, and establish the user behavior mode decision tree, on the basis of complete analysis of cloud computing smart home user behavior patterns mining method design. Through the comparison experiment with the traditional method, it is concluded that the designed mining analysis method based on cloud computing has higher robustness, the proposed cloud computing-based intelligent home user behavior pattern mining method has good application space.

Aiming at the problem that the existing behavior recognition method can not extract the human body interaction area, resulting in low recognition rate, a behavior recognition method based on machine learning and Fisher vector coding is proposed. Constructing artificial neural network based on machine learning, designing the main steps of backward propagation neural network, making the cost function minimum; using the depth continuity of the image to extract the foreground part of the video motion, multiplying with the corresponding 2D video frame to detect the time domain motion Behavior; Solving the dual quadratic programming problem of Fisher support vector machine, obtaining its optimal solution and completing behavior learning; segmenting the current frame image, solving the normal vector to extract the moving target, and completing the behavior recognition method based on machine learning and Fisher vector coding the study. In order to verify the effectiveness of the design method, a comparative experiment was designed. The experimental results show that the average recognition accuracy of the design method is 7.6% higher than the traditional method.

Aiming at the problem that the traditional scheduling method can’t deal with a large number of data quickly when dealing with social network resource data, a new scheduling method of social network resource data based on multi-Agent technology is proposed. Firstly, the social network scheduling framework is designed, using the two-level structure of Agent and three CDN management domains to hide the distribution and heterogeneity of different resources, setting the upper limit trigger conditions of data in each management domain of the framework, using reasoning tools to infer and calculate the SLA comprehensive level of each network operation node, calculating the proportion difference between various resources, selecting the appropriate bias a two-stage resource scheduling method is used to realize resource data scheduling. The experimental results show that: compared with the traditional scheduling method, the social network resource data scheduling method based on multi-Agent technology can maintain the processing time in about 10 s with the increase of data volume, and the processing time is shorter, which is more suitable for practical use.

The recognition rate of traditional face recognition algorithm to the face image with occlusion is not high, resulting in poor recognition effect. Therefore, this paper proposes a partial occlusion face recognition algorithm based on recurrent neural network. According to the different light sources, the high filtering function is used to analyze the halo effect of the image, realize the preprocessing of partially occluded face image, set up the global face feature area and the local face feature area according to the image features, and extract the global and local features of the image; based on the time and structure features of the recursive neural network, establish the local subspace, and realize the local face image recognition Law. The experimental results show that: compared with the traditional algorithm, the face recognition algorithm studied in this paper has a higher recognition rate, and can accurately recognize the partially occluded face image, which meets the basic requirements of the current face image recognition.

In order to improve the effective ability of sports, it is necessary to carry out sports timing training, construct the wireless communication network of sports timing training, and propose a sports timing training method based on wireless communication. A training model of motion timing in wireless communication network based on spatial interval equilibrium regulation and piecewise load distribution is constructed. The wireless communication transmission channel model of motion timing training is constructed, the channel adaptive equilibrium allocation of motion timing training in wireless communication network is carried out by using spatial equilibrium scheduling method, the orthogonal matching signal tracking model of motion timing training is established. Combined with the load block equilibrium allocation method of multiplex motion timing training, the optimal allocation of motion timing training is carried out. the effectiveness of motion timing training information transmission in wireless communication network is characterized by the number of endpoint paths, and the anti-interference ability of communication is improved by combining the interference suppression method of training environment. The simulation results show that the output quality of motion timing training in wireless communication network is high and the bit error rate (BER) is low, which improves the balanced distribution ability of motion timing training load in wireless communication network.

By the localization and recognition of human moving target in video image, combined with the information of human motion feature in video image, the moving target localization and visual reconstruction is realized, this paper analyzes the feature quantity of moving objects in video image, improves the training level, and proposes a moving objects positioning technology of video image based on computer vision and 3D feature point reconstruction. According to the moving feature position of human body, the 3D information modeling and image acquisition of moving target is carried out by using video information acquisition and spatial feature scanning methods. The moving feature points of the collected moving target video image are calibrated and arranged, and the 3D edge outlines feature point set of human skeleton is extracted and represented as a high dimensional vector to form the regular feature database of moving target video image. The moving points in the regular feature database of moving target video image are fusion to realize the reconstruction of moving target video image and the location of moving target. The simulation results show that the method has good real-time and accuracy in moving target location of video image, has strong ability of 3D marking of human moving points, and has high accuracy of extracting moving motion features.

In order to improve the anti-interference control ability of vacuum packaging machine, the design method of anti-interference control system of vacuum packaging machine based on wireless network is put forward. The system design is divided into anti-interference control algorithm design and hardware design of vacuum packaging machine. The parameter distribution model of anti-interference control of vacuum packaging machine is constructed. The output voltage, power and potential field of vacuum packaging machine are taken as the control constraint parameters. The difference between the output phase current of the vacuum packaging machine is calculated, the saturation function is constructed to analyze the control decision variables of the vacuum packaging machine, the sampling output voltage stabilizing characteristic quantity and PWM duty cycle are taken as the modulation parameters, the equivalent mechanical power and equivalent electromagnetic power of the vacuum packaging machine are calculated, and the network output design of the anti-interference control system of the vacuum packaging machine is carried out in the wireless network. The hardware design of anti-interference control system of vacuum packaging machine is carried out by using FPGA. The test results show that the output stability of anti-interference control of vacuum packaging machine is good and the robustness of motor control is strong.

In order to realize the voltage conversion control of vacuum packaging machine seal, a design method of vacuum packaging machine seal transformer based on MCU control is proposed. The main function modules of the system are vacuum packaging machine seal physical index collection module, integrated control module, upper computer communication module, program control module, transformer load balance control module, human-computer interaction module, etc. In the intelligent auxiliary control system, the program cross compiling design of the sealing voltage conversion control of the vacuum packaging machine is carried out, the voltage conversion control is identified in the sealing process of the vacuum packaging machine, the characteristic quantity of the sealing transformer of the vacuum packaging machine is extracted, the human-computer interaction design of the sealing transformer of the vacuum packaging machine is carried out by using the visual operation and maintenance information management method, and the DSP is integrated The information transmission control of the sealed transformer of the vacuum packaging machine is realized. The simulation results show that the designed sealed transformer of vacuum packaging machine has good reliability and strong adaptive control ability, and improves the sealing pressure control ability of vacuum packaging machine.

In order to improve that segmentation quality of the video image of the marathon, a video image segmentation algorithm based on machine learning is proposed. Constructing the edge contour feature detection and the pixel feature point fusion reconstruction model of the marathon moving video image, carrying out multi-level feature decomposition and gray pixel feature separation of the marathon moving video image, and establishing a visual feature reconstruction model of the marathon moving video image, the feature segmentation and the edge contour feature detection of the marathon moving video image are carried out in combination with the block area template matching method, the similarity information fusion model is used for carrying out the video information fusion awareness and the block area template matching in the process of the marathon moving video image segmentation, the fuzzy feature quantity of the moving video image of the marathon is extracted, and the machine learning method is adopted to realize the fusion awareness and the segmentation quality evaluation of the marathon moving video information. The simulation results show that the method is good in image segmentation quality and high in image recognition, and the output signal-to-noise ratio of the motion feature reconstruction of the marathons moving video is high.

In order to improve the accuracy of long jump in long jump, combined with computer vision image processing method to correct the long jump trajectory in long jump, an adaptive tracking method of long jump trajectory tracking image based on machine vision tracking detection is proposed, and the video point frame scanning method is used to collect the long jump trajectory tracking image. The image of long jump athletes is segmented by adaptive pixel fusion method, and the automatic tracking and recognition of long jumpers’ motion trajectory tracking image is carried out based on dynamic feature segmentation. The grey feature quantity of long jump trajectory tracking image is extracted, and the neighborhood distribution model of long jump in long jump is constructed. According to the dynamic evolution characteristic distribution of the long jump trajectory, the dynamic characteristics of the long jump trajectory are analyzed, and the image segmentation of the long jump track tracking is realized by combining the spatial neighborhood enhancement technology, and the adaptive tracking of the long jump trajectory in the long jump is realized according to the image segmentation results. The simulation results show that this method has high accuracy in adaptive tracking image of long jump athletes, and improves the accuracy of long jump in long jump.

The problem of low accuracy and poor ability of tactical analysis of current tournament analysis strategy, to improve the statistical intelligent analysis ability of the tactical analysis of sports games, it is necessary to carry out the tactical analysis and statistics of the sports games, and put forward a game strategy analysis and statistical model based on a large data association rule mining. firstly, a data acquisition and diagnosis analysis model of a sports game strategy analysis is constructed, and a sports game tactical analysis data information flow model is constructed, and the statistic characteristic quantity in the sports game tactical analysis is extracted, the method of fuzzy correlation fusion is used to analyze the interference of the statistical data in the tactical analysis of the sports game, and a plurality of known interference frequency components are removed. in that method, a large-data association rule mining algorithm is adopted to carry out statistical data identification and characteristic analysis in a sports game and tactical analysis, an optimization characteristic solution of the game and tactical statistical information analysis is established, and a self-adaptive optimization of the game and tactical statistical information analysis is carried out by adopting a simplified gradient algorithm, the feature analysis and data optimization recognition of the game tactics analysis are realized by combining the self-correlation feature matching method. The simulation results show that the method is used to carry out the tactical analysis of sports games and the accuracy of the statistics, so as to improve the ability of the tactical analysis of the sports games, so as to improve the effect of the competition.

In order to improve the ability of the accurate detection and recognition of the swing of the training limb of the sprinter, a method of image recognition based on the visual signal processing for the motion of the limb is proposed. The method comprises the following steps of: analyzing the motion characteristic quantity of the swing of a sprinting training limb, carrying out a sprinting training limb swinging image acquisition by adopting an infrared characteristic scanning technology, carrying out edge contour detection on the collected sprinting training limb swinging image. Carrying out contour segmentation and characteristic identification of the swinging of the limb of the sprinting training limb in combination with the image segmentation technology, constructing a gray histogram distribution structure model of a sprinting training limb swing image, and carrying out effective extraction of the motion characteristics of the limb swing and the motion characteristic of the sprinting training limb by adopting a regional block matching method, The method of multi-dimensional space reconstruction is adopted to simulate the motion of the limb, and the visual signal processing and the key feature point calibration method are adopted to calibrate and detect the characteristic points of the limb swing of the sprinting training, so as to realize the optimal recognition of the swing of the training limb of the sprinter. The simulation results show that the method of the invention has the advantages of high accuracy, good characteristic identification ability and the capability of improving the swing image recognition ability of the sprinter training limb.

In order to improve the secure storage and fault-tolerant ability of intelligent agricultural information remote data storage system, a fault-tolerant method of intelligent agricultural information remote data storage system based on block chain is proposed. Combined with the statistical feature analysis method, the fault-tolerant characteristics of the intelligent agricultural information remote data storage system are analyzed, the correlation characteristics of the intelligent agricultural information remote data are extracted, the block chain control model of the intelligent agricultural information remote data is established, and the block storage and feature matching design of the intelligent agricultural information remote data is carried out by using the autocorrelation feature detection method. The block chain storage structure of intelligent agricultural information remote data security storage is established, combined with the optimized block chain control scheme, the optimal storage structure of intelligent agricultural information remote data is reorganized, the structure reorganization and feature reconstruction of intelligent agricultural information remote data are realized by using fuzzy clustering and vector quantification coding scheme, and the fault tolerant strategy optimization of intelligent agricultural information remote data storage system is realized. The simulation results show that the design of intelligent agricultural information remote data storage system based on this method has good fault tolerance and strong coding and decoding ability, which improves the security of intelligent agricultural information remote data storage and management.

In order to improve the multi-parameter monitoring capability of the intelligent agricultural greenhouse, a multi-parameter monitoring system design scheme of the intelligent agricultural greenhouse is proposed based on the artificial intelligence control, the monitoring system mainly comprises a greenhouse multi-parameter information acquisition module, a smart agriculture greenhouse bus control module, a greenhouse temperature information fusion model, a program loading module, a remote communication module, an embedded scheduling module and a human-computer interaction module, the design interface program realizes the man-machine interaction design of the multi-parameter monitoring system of the intelligent agricultural greenhouse, and the main control module of the multi-parameter monitoring system of the intelligent agricultural greenhouse is constructed, and the embedded development of the intelligent agricultural greenhouse integrated intelligent monitoring system is carried out based on the bus under the IEEE488.2 standard, Combined with the application environment of the multi-parameter monitoring system of the intelligent agricultural greenhouse, the self-adaptive output conversion control of the intelligent agricultural greenhouse is carried out, and the optimization design of the multi-parameter monitoring system of the intelligent agricultural greenhouse is realized under the DSP environment. The test results show that the method can be used to monitor the multi-parameter of the intelligent agricultural greenhouse, and the stability of the system is strong.

Under the environment of Internet of things, agricultural network information service is open and resource sharing. In order to improve the intelligence of agricultural network information service under the environment of Internet of things, an agricultural network information resource sharing system based on Internet of things is constructed. The overall design description and function modularization analysis of agricultural network information resource sharing system are carried out. The system design includes agricultural network information service resource retrieval module, agricultural network information resource integration processing module, bus control module, resource information fusion module, program loading and compilation module and human-computer interaction module. The bottom module of agricultural network information resource sharing system is designed by using B/S architecture protocol and bus server system, the retrieval of massive agricultural network information service resources is designed based on Internet of things technology, the information dispatching network center of agricultural network information service resources is established under the environment of Internet of things technology, and the Internet of things networking design of agricultural network information resource sharing system is carried out by using network networking methods such as ZigBee and GPRS. The process management and file configuration are carried out under MVB bus control protocol, and the software development and design of agricultural network information resource sharing system are realized under embedded ARM environment. The test results show that the information resource sharing system of agricultural network based on Internet of things technology has good human-computer interaction and resource scheduling, and the execution time cost is small and the reliability is high.

In order to improve the ability of optimal allocation of agricultural information resources, analysis of agricultural information resources allocation efficiency, optimize the allocation of agricultural information resources, has become an important issue. An evaluation method of allocation efficiency of agricultural information resources based on huge data is raised, and the software development and design of configuration efficiency evaluation model are carried out in combination with embedded LINUX system. The big data distributed storage structure model of agricultural information resources is constructed. Quantitative regression analysis and adaptive game method are used for quantitative evaluation of agricultural information resources, big data information of agricultural information resources is excavated, regional grid clustering method is used for classification and recognition of agricultural information resources, and information fusion and adaptive scheduling of agricultural information resources are carried out under the optimized information fusion model. Embedded Linux technology is used to develop the evaluation model of agricultural information resource allocation efficiency in C/S client. The system includes data processing module, agricultural information resource allocation module, program loading module, bus scheduling module and human-computer interaction module. The integrated design technology is used to realize the software design of agricultural information resource allocation efficiency evaluation model. The test results show that the designed evaluation model of agricultural information resource allocation efficiency has good reliability and strong human-computer interaction ability, which improves the quantitative evaluation and optimal allocation ability of agricultural information resources.

In view of the design of seamless and fast panoramic image generation system, the traditional panoramic image generation system has the problems of poor image generation effect and long calculation time. The design of seamless and fast panoramic image generation system based on VR technology is proposed. According to the image edge processing mechanism, the hardware structure of the system is designed, and the software function of the image processor in the hardware is designed. The process of image edge recognition is described in detail, and the edge is tracked and refined to accurately track the direction and offset of the pixel. The boundary pixel information is recorded by VR technology, and the seamless system of panoramic image is realized by combining the corresponding image data. The experimental results show that the highest generation effect of the system is 90%, which shortens the calculation time and provides effective support for the research of automatic image generation system.

The existing technical means cannot effectively determine the negative impact of horizontal and vertical data behavioral risk on the return of enterprise financing investment. In order to solve this problem, the behavior characteristics of enterprise financing investment risk data were analyzed. Through determining the subset of risk factors and selecting the redundancy degree of SMT (Securities Margin Trading) index, the risk coefficient evaluation of financing investment was completed. On this basis, through the calculation of investment risk data time expenditure behavior and space expenditure behavior, determine the constraints between various data behavior characteristics, complete the enterprise financing investment risk data behavior characteristics analysis. The results show that the negative influence of horizontal and vertical data behavioral risk on enterprise financing investment is effectively restrained after applying the new investment risk behavior analysis method.

In order to improve the sharing ability of network database in the visual environment of learning behavior big data, a design method of network database sharing platform based on big data visual analysis of learning behavior is proposed. The piecewise linear coding method is used to collect the characteristics of the network database in the visual environment of learning behavior. The characteristics of association rules of network database shared resources in big data visual environment are extracted, the optimal structure of network database shared resources is reorganized by adaptive spatial resource reorganization method, the optimal storage spatial distribution model of network database shared resources is established, and the statistical analysis method is used to search the association of network database shared resources in big data visual environment. The feature quantity of association rules of network database sharing resources is extracted, and security sharing of network database sharing resources is carried out according to the clustering of feature quantity. The network database sharing platform is designed in embedding ARM environment. The simulation results show that this method has good adaptability and security, and improves the level of resource sharing.

In order to improve the security of the distributed optical fiber sensing network, the self-adaptive detection of the fiber sensing network needs to be carried out, and an overlap detection algorithm under the APT attack of the distributed optical fiber sensing network based on the spectral characteristic component and the big data analysis is proposed. the large data sampling model of the network APT attack is constructed, the attack characteristics and the related properties of the distributed optical fiber sensing network virus are simulated by adopting the spectrum correlation characteristic detection and the large-data quantization characteristic coding, and the large-data fusion and feature extraction of the APT attack information are realized, the output abnormal characteristic detection of the distributed optical fiber sensing network is carried out through the feature extraction result, a distributed optical fiber sensing network intrusion large data statistical analysis model is constructed, and a narrow-band signal spectrum offset correction method is adopted, And calculating the connection probability density and the individual infection probability of the APT attack node, and improving the detection capability of the network APT attack. The simulation results show that the algorithm can effectively implement the network APT attack detection, improve the security detection capability of the network APT attack, and has a good network security protection capability.

In order to achieve high precision localization, this paper presents an integrated localization scheme employs two particle filters (PFs) for fusing the inertial navigation systems (INS)-based and the light detection and ranging (LiDAR)-based data. A novel data fusion model is designed, which considers the robot position error, velocity error, and the orientation error. Meanwhile, two-PFs based data fusion filer is designed. The position errors measured by the two-PFs in real tests is 0.059 m. The experimental results verify the effectiveness of two-PFs method proposed in reducing the mobile robot’s position error compared with the two-EKF method.

Aiming at the problem that the drawing of ceramic products designed and produced by the original system is low in definition, which leads to poor picture effect, this paper proposes the design of parametric CAD system of ceramic products based on virtual reality technology. Through sensor-mcu control system, complete data collection, build a computer control system platform, select the appropriate data transmission port, complete data transmission; Application of virtual reality technology, and through the data query, modify, and 3 d parametric regeneration to complete its structure system design work process, improve the search efficiency of menu file, according to user requirements, remove or add a menu item, calculating weights of the Bezier curve, design the user interface and high efficiency, adjust the memory image file, complete the system design. The comparison experiment is designed in CAD software, and the image effect of the system in this paper is compared with that of the traditional system. The ceramic products produced by the design system have higher clarity and better effect.

In view of the narrow data source range of traditional residual oil distribution numerical simulation, a numerical simulation method of residual oil distribution based on oilfield data analysis is studied. Firstly, the principle of big data was put forward to complete the collection and analysis of oilfield information. Then, the three-dimensional simulation model of oilfield was designed by haiyan software. Then, the grid structure of oilfield saturation setting model was divided. By comparing the range of data sources, it is verified that the proposed numerical simulation method can effectively extend the range of data sources and improve the accuracy of numerical simulation.

CT image diagnosis technology had developed rapidly in China. In clinical testing, large-scale data gradually existed in the form of sequences. Data clustering was the integration of different substances in an image according to certain properties, but there were still many problems in the use of commonly used data clustering methods in medicine. The current sequencing big data clustering method analysis was still in the research stage and had very important significance. This paper proposed a large-scale batch clustering method based on multi-frame CT images, which was compared with the traditional clustering method, and hoped to provide assistance for clinical applications.

In the research of scale space fusion of medical big data video image, due to the different spatial management methods of multi-source video image, the efficiency of multi-dimensional scale space fusion is low, so a new image scale space fusion method is designed according to the characteristics of medical big data video image. From the aspect of video image visualization and multi-dimensional space fusion, the space fusion method is designed respectively, and the sampling survey method is adopted for experimental analysis to obtain experimental data. The experimental results show that the proposed method is feasible and reliable.

Aiming at the problem of high false alarm rate and failure rate in traditional data mining methods of malicious attacks, a data mining method of malicious attacks based on characteristic frequency is designed. Preprocess the original data in the data set, select the minimum attribute subset, use the discretization to process the unified data format, take the new subset as the input of feature frequency extraction of malicious attack data, extract the feature frequency according to the different protocols of malicious attack data transmission, integrate it into the value data mining algorithm, and use the spatial mapping principle to realize the malicious attack data excavate. The experimental results show that: compared with the traditional data mining method, the false alarm rate and failure rate of the designed malicious attack data mining method based on the feature frequency are reduced by 0.3 and 0.2 respectively, which shows that the method is more suitable for practical projects.

In order to solve the problem that the traditional image recognition algorithm can not guarantee a relatively stable recognition accuracy and poor robustness under a variety of interferences, the image recognition algorithm based on depth level features is studied. After preprocessing, such as filtering and enhancing, the image to be recognized is segmented. The segmented image is input into convolution neural network, and the feature of depth level is extracted from the neural network. The feature points of the extracted deep level features are matched to realize image recognition, and the algorithm of image recognition based on the deep level features is designed. Compared with the traditional image recognition algorithm, the designed image recognition algorithm can ensure a more stable recognition accuracy and has better robustness.

In the process of the development of science and technology in China, in order to maximize the use of network and high-tech resources, more and more fields are increasingly being used for artificial intelligence during this period to meet the growing demand for high technology and requirements. Artificial intelligence is necessary to detect the shadow of infrared remote sensing images. Therefore, in the field of image detection, artificial intelligence must be applied appropriately. In this paper, through the research on the characteristics of artificial intelligence and the artificial intelligence infrared shadow image detection method, the importance of artificial intelligence for infrared remote sensing image detection is deeply analyzed. The significance of research on infrared remote sensing image shadow detection based on artificial intelligence is emphasized.

The rapid development of computers has improved the scope of dynamic knowledge network learning applications. Online learning has brought convenience to people in time and place. At the same time, people began to pay attention to the efficiency and quality of online learning. At present, the network knowledge storage system is distributed storage system. The distributed storage system has great performance in terms of capacity, scalability, and parallelism. However, its storage node is inexpensive, and the reliability is not high, and it is prone to fault. Based on the designed fault detection model detection path, relying on building the knowledge data node fault detection mode, constructing the knowledge data link fault detection mode, completing the fault detection model detection mode, and finally realizing the dynamic knowledge network learning system fault intelligent detection technology research. The experiment proves that the dynamic knowledge network learning system fault intelligent detection technology designed in this paper reduces the fault rate of the network learning system by 37.5%.

Because the latitude and longitude feedback of the traditional ocean monitoring buoy is not accurate, a method of ocean monitoring buoy positioning based on visual feature recognition is proposed. By selecting buoy shape and material, the communication module and image acquisition module are designed Complete hardware design. In the software design, according to the angle between the lowest pixel, the highest pixel and the horizontal direction, the positioning analog signal is selected. Through the E-R model and the buoy information table field, the database design is completed. So far, the design of the buoy positioning system based on visual feature recognition is completed. According to the simulation results, the longitude and latitude coordinates of the design system feedback can be accurate to four decimal places, which shows that the designed positioning system is more accurate.

In view of the problem of poor real-time transmission of the traditional information transmission method of marine monitoring buoy, the transmission speed is slow. Based on this, iridium data communication system is applied to the information transmission of ocean monitoring buoy. Through the establishment of the overall framework of Iridium satellite communication system, determine the buoy information transmission network protocol. On the basis of the network protocol, use Iridium satellite data communication system to obtain the location information of the marine monitoring buoy, and then process the obtained information. Finally, upload the processed buoy information to complete the transmission of the marine monitoring buoy information. Compared with the traditional method of information transmission of ocean monitoring buoy, the experimental results show that the method of information transmission of ocean monitoring buoy using iridium data communication system can complete the transmission of buoy information in a shorter transmission time, with better real-time transmission.

In view of the low virtual force coverage of traditional wireless sensor networks. A virtual force coverage control system based on wireless sensor network is designed. Hardware design mainly includes network interface, processor, control chip and network coordinator. In the software part of the system, firstly, the virtual force coverage control node is selected. On this basis. The optimal control of virtual force and coverage of wireless sensor networks. The virtual force coverage control system of wireless sensor network is designed under the background of big data. The experimental results show that. Under the background of big data, the coverage of virtual force control system of wireless sensor network is higher than that of traditional system. It has a certain practical significance.