Design And Implementation Of Monitoring Software For Elderly People Living Alone Based On Video Analysis

With the problem of aged tendency of population and empty-nest aggravating in our country,the issue of guardianship for elderly people living alone has received widespread attention.However,due to the increase of social pressure,most of the children are away from the elderly and employed in other places,and they cannot monitor the elderly in real time.In recent years,the rapid development of science and technology has brought a new way to the elderly care.Many researchers have done a lot of work and made great progress.Some monitoring software products have also been applied in life.There are still some shortcomings,mainly manifested in the weak ability to analyze abnormal behaviors of the elderly,and the existence of single software function design and inability to provide diversified elderly care services.Based on this,this thesis uses video analysis technology as the core to design and implement a monitoring software system for intelligent identification and early warning of indoor abnormal behaviors of elderly people living alone.At the same time,it realizes the combination of family care and community services to provide diversified health care for the elderly service.Specifically,this thesis contains the following work:1.This thesis summarizes the research background and development status at home and abroad,and analyzes the existing problems in the current work.Based on this,the overall analysis of the monitoring software for the elderly living alone based on video analysis is carried out.By analyzing the needs of different users of the software,the overall structure of the software is clarified,and a software development plan based on the combination of C/S and B/S is proposed.2.In view of the weak ability of existing monitoring software to analyze the abnormal behavior of the elderly,this thesis takes the video analysis technology as the core and researches the elderly behavior recognition algorithm based on the video analysis.This thesis first uses the improved visual background extraction algorithm to extract moving foreground targets;Then,it extracts four kinds of characteristic parameters: the invariant moment of Hu,the minimum external moment aspect ratio,the ellipse fitting features,and the speed of movement to characterize the behavior of the elderly;Finally,the selected feature parameters are combined to train the support vector machine classifier model,and the model is used to identify the behavioral posture of theelderly.Experiments show that the algorithm is selected in this thesis meets the expected needs and provides theoretical support for the design of monitoring software.3.According to the demand analysis and algorithm research,this thesis has designed and implemented the software function module and database,including sub-modules such as abnormal behavior identification,system management,remote monitoring,emergency rescue,and visit care.4.The monitoring software is tested and the results are analyzed.The functional and non-functional tests are carried out on the monitoring software by analyzing and selecting test plans,setting up a test environment,and designing test cases,and the test results are analyzed.The test results show that the monitoring software designed and implemented in this paper has achieved the expected goal.