The Research And Application Of Gesture Recognition Method Based On Android OS

In fields of intelligent mobile phone, vehicle system, tablet computer and digitaltelevision, the Android operating system has occupied a considerable market share. TheAndroid with the characteristics of stable, open and free is recognized by the majormanufacturers. At present, the man-machine interactive of intelligent terminal based onAndroid is mainly by screen touch technology, and by some physical keys. Some products arealso equipped with a trackball, touchpad and other auxiliary control method. However, thecurrent way of human-computer interaction is still has some limitations, this limitation ismainly reflected in the human-computer interaction distance.In order to break through the distance constraints of human-computer interaction, andallowing users to interact with the device with more natural way, this paper studies a methodthat can control Android equipment with contactless natural gestures. The focus of the studyis to be a computer vision library OpenCV transplant to the Android platform, and on this toachieve a natural gesture recognition system that can develop for two times.This paper introduced background and architecture theory of the Android system andOpenCV library, the current man-machine interactive mode in Android equipment, discussedin detail how to transplant and deployed the computer vision library OpenCV to the Androidsystem, which the OpenCV library is the basis of image processing library to recognizingnatural gesture information processing. Focuses on the framework design of the naturalgesture recognition system, as well as the system design and implementation method of eachmodule. The paper describes the development method that can be developed for two times,and including the addition of classifier, customizing decision command and event processingaction. This can help other developments to achieve the desired application functions.Finally, done test for the gesture recognition system. This system can recognize statichand gesture based the classifier, and then make the appropriate decision. The test resultsindicated that this system meets the real-time processing need, and the movement is stable.This interaction is more consistent with the user’s natural behavior. In the interactive process,it takes users through gestures to transmit commands to the Android smart device. With thenon-contact human-computer interaction technology continues to mature and improve, it willbecome a main control mode to Android intelligent device.