Design And Implementation Of Portable Motion Detector

Appropriate exercise is an essential part of healthy lifestyle.In many areas, accurate human motion pattern recognition and counting steps have its significance. By the way of monitoring, analysis and identification of human motor behavior motion detector can provide all kinds of health information and health programs and make people’s lives healthier. With the mobile Internet is increasingly integrated into our lives, it will become one of the new opportunity, a variety of new mobile Internet applications, including fitness, personal navigation, social networking, etc. has been more and more attention.In this paper, the research objectives is motion detection implemented. Based on intelligent motion detector health, design and implement a portable device to meet people’s movement for accurate monitoring needs.Existing health facilities have some problems, for example, sport mode judging and for non-normal state monitoring data is not precise enough. For these two issues this paper using the classification algorithm and adaptive peak detection algorithm to make improvements.Classification algorithm is based on analysis and processing of the extraction of the acceleration sensor signal. First, preprocessing the data, this includes three steps noise filter to remove interference gravity, windowing processing and analysis.After to selecting the characteristic of processed data and using a classifier to distinguish and distinguish different patterns of movement according to the characteristic values.Now common pedometer algorithm include peak detection algorithm and autocorrelation analysis algorithm. These two methods walking pedometer accuracy can reach more than 98% during normal state. But in running, stairs and other non-state mode of its normal pedometer accuracy can only reach about 60%.The main reason for this phenomenon is that the above two methods in the human body when the pedometer is a constant determined threshold value. In fact, In fact the body is walking, running, up and down stairs and other different sport mode acceleration determination threshold value which has a considerable gap. To solve this problem this paper proposes an adaptive peak detection algorithm, it is according to the different of normal state and no-normal state, using the corresponding threshold interval discriminating different movement patterns, fix the size of the neighborhood window and comparison the field of adaptive to achieve accurate pedometer.