| At present,the population development presents the trend characteristic of Shaozi and aging,and the degree of population aging is further deepened.However,with the growth of age,the self-care ability and physical function of elderly people living alone continue to decline,so even in daily home life,it is inevitable that various safety accidents will occur,such as gas leakage,burglary and so on.In addition,falls are one of the key causes of disability and death in the elderly,and the proportion of falls caused by factors in the home environment is very high.Therefore,this paper aims to design a smart home system for the elderly living alone at home,aiming to improve their sense of happiness and security in home life.The system is mainly divided into three core parts: indoor environment data acquisition,fall detection based on Yolov5 algorithm and Qt human-computer interaction interface.(1)Indoor environment data collection is to simulate the actual home life scene.STM32F103C8T6 single chip is selected as the main control chip for the environmental parameter collection of the terminal node and the subsequent linkage work,and then the collected indoor environment parameters are sent to the CC2530 module of the terminal node by means of serial communication.Finally,terminal nodes and coordinators distributed in different indoor scenes are simulated to complete indoor Zigbee wireless sensor networking.(2)Based on the Yolov5 algorithm,partial Fall Detection Databases(FDD)were integrated with indoor,outdoor and multi-scene fall datasets selected through Internet search,and the images that fell or the body began to lean were marked as "fall-down".Then,a total of 12,746 fall data sets were divided into training sets and verification sets with a ratio of 8:2,and model training was carried out.Finally,a Yolov5 fall detection model with an average detection accuracy of 91% was obtained.(3)In the Windows environment,the functions of the first two parts are integrated and expanded through Qt man-machine interface.On the one hand,the interface can display the indoor environmental parameters in real time,so that the elderly can know the indoor environment;On the other hand,users can choose whether to enable the fall monitoring function according to their personal wishes.When a fall event is detected,the system will decide whether to send the alarm message to the mobile phone of the specified object according to the preset priority based on the actual situation and user feedback.At the same time,the elderly health information file database is set up,so that the rescue personnel can refer to the information in this database to better rescue.Finally,the node function test,fall effect and alarm test of the system are completed.The results show that the system has good functional effect.The system designed and developed in this paper can provide a more comfortable,happy and safe smart home life for the elderly living alone without increasing any burden,especially the fall monitoring function and the follow-up linkage alarm setting.It has certain reference value and practical significance for solving the daily home life safety problems of the elderly living alone,which are widely concerned by the society. |
PDF Full Text Request