Research On Android Face-tracking Application In The Intelligent Visual Internet Of Things

The Internet of Things is a kind of intelligent networks which be used to locate, track, identify and supervise the people and objects. The concept of Three-Network Convergence promotes the rapid development of the Internet of things. Intelligent visual internet of things is the visual perception part of the Internet of things. With the knowledges of Communication, Computer Network, Image Processing, Software Development and other disciplines, this paper researches the Android Face-tracking application in the Intelligent Visual Internet of Things.This paper introduces the three parts of the Intelligent Vision Internet of Things-Sensing Layer, Network Layer and Application Layer, involving several techniques of the Information field. For transmitting information wireless with the Network Layer, Wi-Fi wireless network cameras-IPCAM are utilized in the Sensing Layer as the node to capture video data. The Network Layer takes advantage of the Google Drive in which the video data are stored. Users can access their data through the network anytime and anywhere. Building the Sensing Layer and Network Layer realizes wireless communication and cloud storage, which not only provides convenience for users, but also improves the security of data.In this paper Face-tracking technique uses CAMSHIFT algorithm, which is a non-parametric iterative continuous adaptive algorithm. The color histogram of target area is backprojected for tracking. It is simple and high real-time. The paper introduces the face detector based on Adaboost algorithm, H, S and V components of HSV color space and simple motion estimation information into the traditional CAMSHIFT algorithm, estimation information, not only to achieve face-tracking automatically, but also to improve the accuracy and real-time of face-tracking.Finally, for compiling Android Face-tracking application, with the assist of OpenCV function library, this paper takes advantage of the new CAMSHIFT tracking algorithm, shares the dynamic library compiled by the Android NDK, and then uses the JNI to integrate the Java code and native code. The Application Layer of Intelligent Face-tracking Internet of Things is made up of intelligent mobile terminal equipment and Android Face-tracking application software, in which the users realize intelligent applications of this system. Experimental results show that this application can achieve better tracking effect.