The Research Of Achieving Object Tracking In Android Platform Based On OpenCV

The moving target tracking technology is an important field in machine vision. With thefurther research, the target tracking is used in military guidance, visual navigation, robot,intelligent transportation, public safety. Because of the electronic devices performancecontinuously improving, target tracking application is more convenient and smart. Therefore, aapplication is designed to achieve tracking object based on Android mobile platform, andimprove the moving target detection method.Firstly, the paper analyses the present status of object tracking technology, introduce theprinciple and structure of Android system and OpenCV library.After, the method that how to translate and load the OpenCV library into the Androidsystem is discussed in detail. OpenCV vision library is a function library as a basic imageprocessing method. OpenCV manager can call program and function. The design andimplementation of the various components of the system module is discussed in detail. Imagecapture module achieves initializing the camera. Feature calculating module achievescalculating the target center coordinates and the contour. Tracking module achieves trackingthe target trajectory.Finally, by comparing the optical flow method,background subtraction and other method,it uses frame difference method to detect the target area. Since the method need to be setthreshold artificially, when the background lighting changes, the unique threshold can notmeet the requirements.Therefore, it uses the Otsu method to compensate the lack of algorithms.It can follow the changes of the environment and set the threshold adaptive.It uses theCamshift algorithms to capture the moving target, according to the different value of the centercoordinates, it can be observed visually the effect.Then, debug the program on the Androidsystem phone, the regular and irregular shape object can achieve the tracking effect.