The Research And Implementation Of Methods Of Statistics Of The Number Of People And Indoor Positioning Of People In Building Energy Saving Application

Nowadays, along with the dramatically fast development of society and economy,the requirement of the natural resources has been increased rapidly during the dailylife and production of human beings so that how to use the natural resourcesreasonably and keep the environment out of pollution has been raised as a hot topicwhich is valuable to study and research. In the recent years, many scholars andresearchers have devoted themselves into building saving study, The techniqueimprovement of computer, Internet of things and sensors make that how to obtain thenumber of people and the positions of people in a specified area become an importantresearch aspect in building energy saving study.This paper is to provide a real-time indoor population statistic and peoplepositioning solution which is used for building energy saving through net-cameras,RFID (Radio Frequency Identification) devices and network which is constructed byZigBee and Wifi and to briefly describe how these two elements are used in buildingenergy saving application. After investigation and research of video monitoring,pedestrian recognition, RFID locating algorithm andby considering the real-timerequirement and accuracy of population statistic at the same time, this paperintegratesand improves the popular motion detection and silhouette tracking methodto counting people in a specified area and uses the improved Landmarcalgorithm forRFID individual locating. This paper mainly studied motion detection, pedestrianrecognition, silhouette tracking and individual positioning and the research resultshave been described below:1. Motion area detection based on random theory. In this paper, we create abackground model which is suitable for indoor environment through the first threeframes in a video based on random theory and consequence principle of images toobtain a highly accurate silhouette of the motion area from the video frames rapidly.2. Pedestrian recognition based on fuzzy theory. Based on researching andcontracting current methods of pedestrian recognition, and by considering the real-time requirement condition, we propose a pedestrian recognition method byintegrating skin color detection, human body ratio detection, constraining of humanwindow size and creating human-similar element based on fuzzy theory.3. Silhouette Tracking based on optical flow and statistics of the number of peoplebased on layer object. An important aspect of this paper is how to obtain the pathwhich the pedestrian walks, track pedestrian’s motion and do the statistic when thepedestrian has entered another special area. In this paper, we use manual specialbackground area designating and silhouette tracking based on optical flow toimplement the goal that to track pedestrian’s motion, and, moreover, propose amethod based on layer object theory to do the statistic of the number of people whichcan improve the accuracy.4. Individual locating based on RFID devices. In order to get the individual’spositions information in building energy saving application, we use the RFID devicesto recognize and locate the moving object with RF tags. In this paper, we improve theoriginal Landmarc algorithm by add self-adaptive value to increase the accuracy oflocating. And, at last, we test the performance of the algorithm through simulationexperiments.5. At the end of this paper, we propose animplementation solution of a systembased on the statistics of the number of people and individual positioning, brieflydescribe that how the statistic information and position information use in a big spacearea of energy saving application and discuss the potential of energy saving in thefuture.