Mobile Application GUI Similarity Comparison Based on Perceptual Hash for Automated Robot Testing

Abstract

Due to the rapid growth of mobile applications, the requirements for software testing speed are getting higher and higher. Therefore, automated testing is gradually replacing inefficient manual testing methods. However, most of the existing mobile applications are not open source code, and testing open source mobile applications is time-consuming and inefficient. Therefore, we propose automated robot testing based on black box testing. Firstly, YOLOV3 algorithm is used to obtain the information of mobile application interface components. Then, the perceptual hashing algorithm is used to identify the isomorphic GUI of mobile application, and the nodes that are the isomorphic GUI after the jump are merged to improve the test efficiency by simplifying the model. The experimental results show that the method can identify the isomorphic GUI quickly and accurately, and can better simplify the interface model and improve the test efficiency.