Design And Research Of Gesture Recognition And Image Segmentation On Android Operating System

As a major innovation in the field of human-computer interaction， themulti-touch technology，which is rising in recent years，is a computer multiple inputtechnology. With the continuous development of intelligence, more and more smartphones tend to be humane, scientific, and the most representative feature is thesubstation of finger touch for the on-screen keyboard input. Thus, the multi-touchtechnology is brought into full play in intelligent machines. With the support ofinnovative software and large-screen multi-touch, this brand new user interface iseasy for users to control everything with fingers： such as slide view, click to select,double-click to enter, drag and drop, zoom in and out.With its immeasurable potential applications in the field of technologicaldevelopment and social services, multi-point touch technology has providedcustomers with good user experience （UE）. And related technologies of multi-pointtouch have already emerged as a fast-growing industry with promising prospects andthe commercial value should not be overlooked. One of multi-touch technologyprograms is based on the visual technology of the mobile terminals. With thecharacteristics of low cost and easy to extend，the program is especially suited for theapplication of newly-developed interactive platforms such as Twitter.This article tries to improve the system to accurately sense the two-dimensionalcoordinates and pressure of multi-touch point and achieve the cropping and repairfunction of image lasso mainly through the implementation of research on multi-touchsignal detection platforms and cropping gestures match as well as C language andJava language to improve image processing software.Due to the increasingly extensive use of multi-touch applications on mobileterminals, the demand on multi-touch precision is constantly on the rise. The previoussimple touch browsers or zoom function can no longer meet the needs of therapidly-updated interactive media such as Twitter. People have higher requirementsfor real time image processing. Thus, it is of great necessity and practicality todevelop an application with functions of real time image repair based on the featuresof the Android platform and research on image segmentation. In addition, it is also anirresistable to study on multi-touch image processing （such as more complex andaccurate color, graffiti, etc.） across the board. The article mainly focuses on therecognition method of using gesture for image cropping and repair on the Androidplatform.