The Design And Implementation Of Care System For The Elderly Based On Video Image Analysis

Our country is the most populous nation in the world.People's average life is longer and longer with the development of social economy,the progress of technology and the rise of medical level in the 21 st century.China acts as a member of the aging society,and the elderly health and safety issue has caused the attention of people from all walks of life.There is a special group in the elderly—empty nest elderly,who live alone for a long time and their safety problem is particularly prominent.So developing a service system which can monitor the safety and ensure the life privacy of the elderly is an urgent problem in the elderly care area of smart home,and it is also the main research content of this project.Firstly,this paper introduced the development status,the future prospects and related technologies of the subject at home and abroad.Then,according to the system functional requirements and the actual needs of the project,this paper presentedchose the software and hardware architecture and the framework of video image algorithm of the elderly care system which is based on video image analysis.In this paper,I chose STM32F103C8T6 as the core processor of monitoring boards,which equiped with sound sensors,light sensors,infrared pyroelectric sensors and other peripherals.They are responsible for the direct collection of the life information of the elderly,information transmission to the main control board through the SI4432 wireless communication module.I chose OMAPL138 as the core processor of the main control board which equiped with GPRS communication module,Ethernet communication interface,wireless communication module and other peripherals.The main control board was mainly responsible for the elderly life information acquisition,abnormal judgment and communication with the server.The highlight is that the system use the video image algorithm to identify the elderly fall behavior.It detected movement pedestrian by background deduction algorithm based on improved background modeling,removes shadow by shadow removal algorithm based on HSV color space and identified fall behavior by fall behavior feature extraction algorithm.The elderly care system designed in this paper has already achieved basic function after one year's development and debugging.It can be sensitively detect the movement of the human body and achieve SMS alarm if the human fall.Moreover,it can accurately detect the state out of home,toilet exception,getting up,sleep exception and so on of the elderly.The false detection rate of the entire system is tested,which remains at 5% or less.So the expected technical specifications of this system is completed in a good way.From point of view of user,the product does not need to install in the elderly clothing.And the server client is provided with user rights whose privacy protection is better and the cost is low.It meets the user experience in a better way.