Research On Indoor Outdoor Recognition Technology Based On User Behavior Patterns

Nowadays a growing number of mobile applications are centered on the user’s location. The location information not only includes physical coordinates but also logical locations. Indoor or outdoor is an important primitive logical location information, it can be used by many location-based and context-aware applications. Many applications can easily get accurate and effective information from indoor or outdoor scenes. However, these works are based on the premise that the target user’s indoor outdoor state is known, and sometimes such assumption hardly holds in practice. So the research of indoor outdoor recognition, especially whose accuracy and pervasiveness, still has a long way to go.In this paper, we carry on a thorough analysis and study on users’ different behavior patterns of indoor and outdoor, we extract some consistent physical features from user’s motion patterns. And based on these features we present a new indoor outdoor recognition technology. The accuracy and pervasiveness of indoor outdoor recognition are improved by this technology.Firstly, we study the existing indoor outdoor recognition technology, analyzing the application scenarios and pros and cons of each technology. Then we look deep into the recognition technology based on environment sense data. Through this research, we design two new recognition algorithms, including turn frequency algorithm and motion duty ratio algorithm.These two algorithms decrease the influence of environment factors in different circumstances, thus they improve the whole pervasiveness of indoor outdoor recognition technology. Finally, we complete a pervasive, accurate and fast indoor outdoor recognition system on Android platform and evaluate the system comprehensively with data collected from several users’daily life during one month period. The results of accuracy and consumption are satisfying.