Research On Recognition Model Of Hotel Room Guest Based On Multi Data

With the development of Connected Home and the continuous improvement of networking technology,Connected Home has entered all aspects of our lives,Intelligent Hotel is one of them.As the continuous improvement of face recognition technology,the integration of human and certificate is the development direction of Intelligent Hotel.In order to realize the integration of person and certificate,it is necessary to abandon the traditional occupancy mode of using room card.However,compared with unlimited power supply,the function of "inserting cards to get electricity and pulling cards to turn off electricity" of the house card is more in line with the current energy-saving policy.Therefore,we need to study a model that can accurately identify roomers and control hydropower to save hydropower resources.Traditionally,human body induction is mostly used in induction lamp and safety control.The commonly used technologies including infrared human body sensing,microwave human body sensing,camera recognition,etc.However,infrared and microwave technologies have their own short boards,which cannot realize all-round error-free human body identification.Infrared technology and microwave technology have their own short boards,which can not achieve error free human body recognition in all aspects.In order to ensure the privacy of roomers,the camera is not allowed to be installed in the room,so a single human body sensing device can not achieve a full range of human body recognition.In-depth study can effectively realize data processing.However,it takes a lot of time to train and process data,and how to reduce the training and data processing time is also a difficult problem.By analysising of the above problems,the main content of this study is to compare various human body recognition technologies and select the most appropriate multi data fusion technology to build the recognition model for roomers.The human body induction equipment such as infrared and microwave is combined with other pressure sensors to process and judge the data of multiple sensors through multi-data fusion.This study combines the fuzzy theory to improve the DBN(Deep Belief Networks)to achieve multi data fusion,constructs the fuzzy DBN neural network,and optimizes the model by adjusting the boundary parameter.The accuracy rate is increased to 100% and the accuracy rate is 85% by loss recall rate.Comparing the results of the research model with the traditional KNN model,the advantages and practicability of the multi data fusion model are proved.Finally,the SSM(Spring+Spring MVC+Mybaits)framework will be adpoted to build the network platform,and apply the research model to practice.