Research And Implementation Of Indoor And Outdoor Scene Recognition System Based On Smart Phone

In recent years,with the development of mobile Internet and location technology,more and more location-based services applications appear in people's daily life.Although smart phones contain lots of location-based service applications,traditional location-based service applications often divide the service area into indoor scene and outdoor scene.The indoor and outdoor seamless positioning system is a hot spot of people's attention and research,and the basis is accurate indoor and outdoor scene recognition technology.This paper has implemented an indoor and outdoor scene recognition system based on Android smart phone.The main research contents include:(1)The indoor and outdoor scene recognition based on multi-source sensors is researched.First of all,Analyzes the data of various sensor in smart phones in indoor and outdoor performance characteristic.Secondly,Using the light sensor,GNSS module and geomagnetic sensor data,the data of the three sensors were fused,implements the three kinds of indoor and outdoor scene recognition method based on machine learning algorithms.Finally,the recognition accuracy of these methods,running time and scene switching delay are analyzed.The experimental results show that the accuracy of these methods can reach 93% in different experimental scenes,but there is a certain delay in indoor/outdoor scene switching.(2)The indoor and outdoor scene switching based on sequence matching is researched.The indoor and outdoor scene recognition technology based on multi-source sensors has the time delay of scene switching.Therefore,in view of this problem,the performance characteristics of each sensor data when switching between indoor and outdoor are analyzed,and indoor and outdoor scene recognition technology based on the sequence matching is implemented by using the data of light sensor.The experimental results show that these methods have high sensitivity and accurate recognition when switching indoor and outdoor scenes,but their recognition accuracy is affected by more complex indoor environment.(3)Implemented the indoor and outdoor scene recognition system based on Android smart phone.Based on the existing indoor positioning system,an indoor and outdoor scene recognition system based on Android smart phone is implemented by combining two indoor and outdoor scene recognition technologies with alpha-count method.The experimental results show that under different experimental scenarios,the indoor and outdoor scene recognition accuracy can reach 95% in the daytime,and the switching time of indoor and outdoor scenes can reach less than 2 seconds.The indoor and outdoor scenes recognition accuracy can reach 93% in the night,and the switching time of indoor and outdoor scenes can reach less than 3 seconds.And the power consumption of the system is at a low level.This thesis has 48 figures,29 tables,63 references.