Aiming at the problem of multi-sensor image matching when the light intensity is too high or too low, this paper proposes a method of multi-sensor image matching based on joint graph spectrum feature analysis. Firstly, the K-nearest neighbor rule is used to calculate the structural relationship between the corner points in the visible light and infrared images to construct a joint graph. Then, based on Laplacian decomposition, we calculate the eigenvalues of the adjacency matrix in the joint graph to obtain the fea- ture vector of the joint graph, and construct the feature function pair through three-dimensional reconstruction. Thirdly, a maximally stable extremal regions detector based on SUSAN-MSER-SURF is proposed to detect the extreme value position of the feature func- tion pair. Finally, by normalizing the maximum stable extremal regions and matching them, the exact matching results of multi-sen- sor images can be obtained. The experimental results show that the proposed matching method based on the spectral feature analysis of the joint graph can solve the problem of multi-sensor image matching when the light is too high or too low.

Aiming at the problem of power consumption in campus, a smart campus energy consumption monitoring platform scheme based on campus intranet is proposed. The system collects energy consumption data in real time through intelligent data ac- quisition terminal hanging on the Internet, and it is designed based on B/S & C/S. In this paper, the real-time monitoring, statistics, analysis and control of campus energy consumption are proposed. The time series regression analysis method is applied to realize en- ergy consumption prediction and provide decision-making for energy use.

Taking the computer network technology major of Anhui Vocational & Technical College of Forestry as an example, this paper analyzes the purpose and significance of implementing the 1 + X certificate system for the computer network technology ma- jor, and analyzes the current professional development model of our college, to study the direction of the reform of personnel train- ing mode. Through in-depth research, the personnel training objectives of computer network technology major should be clarified, the types of certificates suitable for the development of this major and able to improve the employment rate of students are studied, different professional directions are set up, with emphasis on improving students' practical operation ability and cultivating applied talents. Aiming at the existing problems of computer network technology major in our college, this paper makes a comprehensive analysis and study from the aspects of professional curriculum system structure, teaching resource database construction, teacher strength training, school-enterprise cooperation teaching, practical training and so on, and makes great efforts to construct a 1+X cer- tificate model suitable for the development of computer network technology specialty in our college.

Under the background of engineering education, according to the problems existing in the practice of electrical information students in the college, the teaching reform is carried out from five aspects: combing the practical teaching system, integrating teach- ing, learning and practicing, establishing teachers’workshop, participating in discipline competition and integrating industry with education, so as to improve students' practical ability and achieve good results.

Under the background of engineering education, according to the problems existing in the practice of electrical information students in the college, the teaching reform is carried out from five aspects: combing the practical teaching system, integrating teach- ing, learning and practicing, establishing teachers’workshop, participating in discipline competition and integrating industry with education, so as to improve students' practical ability and achieve good results.

In this paper, a collaborative filtering recommendation algorithm based on user preference is proposed. First of all, the user similarity is calculated according to the length of the longest common subsequence of different user interest sequences and the num- ber of common subsequences, and then the similarity obtained by this algorithm is weighted and mixed with the similarity obtained by traditional collaborative filtering recommendation algorithm. Project recommendation is completed based on mixed similarity and the possible project score by target users is predicted. Finally, by comparing the average absolute error MAE values of three rec- ommendation algorithms in three data sets of Ciao, Flixster and MovieLens 100K, it is proved that the proposed user collaborative filtering recommendation algorithm (XQCF) has improved the accuracy of the recommendation system.

China's ancient architecture has a long history of thousands of years, with its unique art and charm, very popular with the public. With the rapid development of the Internet, the concept of digitization is becoming more and more known. The use of digiti- zation to create an ancient architectural navigation system can play a better role in promoting the tourism landscape. For the digital guide of the ancient architecture of "Nanhua Temple", this paper firstly selects eight main buildings in Nanhua Temple to analyze and study, and completes the three-dimensional modeling, roaming and virtual navigation design, which gives the visitors good visu- al experience and cultural influence.

Under the premise that the remaining electricity is sufficient, the path planning problem of electric vehicles to the target charging pile is the key to the optimization of charging strategies. Normally, when the system is recommended to the user, the sys- tem will make some or a few road sections which have been recommended to different users many times. When a large number of us- ers choose overlapping road sections, they will reduce the current recommendation. The traffic capacity of the road segment makes these charging paths into optimal paths only in a mathematical sense, resulting in local optimization issues. This paper proposes a global path planning method based on load balancing, takes into account the degree of congestion of the section and the user's driv- ing time, to suggest the user a shorter and smooth charging path. Finally, through the simulation experiment, the short-time predic- tion of the traffic flow can reduce the average waiting time of the charging queuing, and the global optimal charging path method based on load balancing can improve the utilization of charging piles and avoid traffic congestion issues.

AccordingtotheactualsituationofNephrologyDepartmentofAffiliatedHospitalofShanxiUniversityofTraditionalChinese medicine. The detailed systematic analysis of the hospital follow-up WeChat Mini Program is carried out, and it is concluded that the hospital doctor follow-up Mini Program should meet the needs of online consultation, doctor's order feedback and other functions. At the same time, the system architecture design and database design of this program are carried out. The system is written through JavaScript language, WXML language and WXSS language, and the development of Mini Programs is finally realized.

Since the curriculum reform, cultivating subject literacy has become the consensus of current front-line educators. When cultivating the core literacy of information technology subjects, teachers often pay attention to the learning process and results, and ignore the cultivation of information awareness. Based on the research on the effective teaching of intelligent robots in elementary schools, it is found that by relying on the cloud classroom platform and using project-based teaching, it can effectively cultivate in- formation awareness in the subject literacy of information technology from three training directions, which are the initiative of ana- lyzing information, the effectiveness of processing information, and the self-learning of sharing information.

CMOS image sensor has been widely used in the field of visible image. It has the characteristics of low cost, low power consumption and high integration. How to transmit the collected image data effectively and improve the transmission distance is one of the main problems in practical application. In addition to the advantages of providing long-distance transmission, Ethernet has the characteristics of high transmission efficiency, simple structure and reliable operation. This paper describes a design scheme of Ether- net transmission system based on CMOS image sensor. Image data is collected by CMOS image sensor, image processing is realized by FPGA, and transmission is carried out through Ethernet interface. The method of displaying image on PC meets the application requirements of the design.

In recent years, the theory of deep learning is booming again. It is widely used in machine learning, visual recognition and auditory recognition. Boltzmann machine is a typical deep learning neural network. There are many training algorithms for its net- work weights, such as contrast dispersion (CD) algorithm, which is classical. However, the current algorithm cannot accurately ob- tain the expected value of network thermal equilibrium state. Only approximate gradient values can be calculated. At the same time, the algorithm has a large amount of computation and a long running time. In this paper, a method of RBM weight calculation is pro- posed. Firstly, RBM is equivalent to Hopfield network. Then the weight matrix is designed by DHNN weight design method. Final- ly, the RBM weight solving problem is transformed into the eigenvalue and eigenvector problem of DHNN weight matrix. An exam- ple is given to illustrate the calculation process and the correctness of the algorithm is verified by the data.

This paper analyzes the demand of data centers (or private clouds), public clouds, 5G Multi-access Edge Cloud and other resources under the trend of digital transformation. The author elaborates the evolvement route of informatization transformation with multi-center distributed deployment for each service and the research on the future network development catalyzed by 5G tech- nology. By analyzing the new "information island chain" problem brought by multi-center (or multi-cloud) deployment, this paper points out the necessity and evolution direction of the integration of cloud and network for telecom operators in 5G era.

Aiming at the problem that the existing IPTV quality monitoring platform cannot match poor quality accounts with user in- formation, and cannot accurately push the details of users with poor quality of broadcast to front-line maintenance personnel, this pa- per implements the IPTV monitor system for poor quality treatment. It uses image recognition algorithms to identify poor-quality ac- counts in batches, automatically extract related information between multiple systems, and directly push the poor-quality informa- tion to front-line personnel according to the maintenance area. The application shows that the system effectively improves the rate of IPTV poor quality in cities and towns, saves labor costs for data extraction, and realizes multi-dimensional monitoring and analysis of IPTV poor quality data.

With the advent of the "Internet +" era, the traditional store's profit model no longer has advantages. As the tangible assets, stores are located in all corners of the country, but the actual revenue that it brings is getting lower and lower. Store passenger flow is decreasing day by day, the effect of in-store preferential activities is not good, it cannot get customers accurately compared with the income, and the cost is higher and higher. The development of the store is in a bottleneck, and the road of reform is imperative. In the context of mixed reform, how to effectively use physical store assets to create new retail models and form new competitive ad- vantages is the focus of thinking. This article focuses on the current status and development trends of the domestic retail industry, and discusses on the current status and future development priorities of China Unicom's physical stores. At the same time, it focuses on "people, goods, field" to discuss on the smart stores transformation road.

The network in cloud environment is different from the traditional network, which has a large amount of data and many ac- cess terminals. Therefore, the requirements of setting security model in cloud environment network are different. This paper propos- es to set up network security model in cloud environment network, using early warning module, adaptive module combined with rule base for intrusion detection. It can greatly improve the security of the network.

According to the performance and functional requirements of blockchain applications, the design concept of integrated baas platform is studied, and the overall architecture of baas supporting the construction of fast alliance business chain is proposed. The design and implementation methods of baas platform are studied from the aspects of high availability computing technology, ap- plication standards and access specifications. Finally, the feasibility of the proposed scheme is verified through three scenarios of us- er login, data link up and data query in 50 blockchains.

This paper takes the protection of user privacy information in express logistics industry as the main research object, and de- signs a scheme to protect user privacy by using RFID technology and multi-layer encryption technology. The proposed scheme can realize the dual privacy protection for both internal and external of logistics companies, and ensure that the responsible person of pri- vacy information disclosure can be reviewed.