A top-down design approach for an automated testing framework

Abstract

Mobile applications have become popular work tools. Portability and ease of Internet connectivity are characteristics that favor this adoption. However, mobile applications sometimes incorrectly process events associated with the user-interaction features. These features include content presentation or navigation. Rotating the devices, and gestures such as scroll or zoom into screens are some examples. There is a need to assess the quality with which mobile applications are processing these user-interaction features in order to improve their performance. In this paper, we present a top-down design approach for an automated testing framework for mobile applications. Our framework integrates digital image processing, GUI information, and historical bug information to identify new bugs based on user-interaction features. Our framework captures images before and after applying the user-interaction features and uses the SURF algorithm to identify interest points in each image. We compared interest points to note differences on the screens before and after applying the user-interaction features. This differences helps to find bugs in mobile applications. The first results show that it is feasible to identify bugs with user-interaction features using the proposed technique.