Intelligent Software Testing based on Visual Feedback

Abstract

GUI testing is an important means to ensure the quality of GUI software. Traditional GUI testing methods often rely on API and have limited scope of use. The introduction of computer vision technology provides a brand-new alternative, based on the image recognition of GUI elements. It can help build visual feedback mechanism to get more information about the target system and control the testing process. In this context, this paper studies intelligent software testing based on visual feedback, designs and builds a three-layer testing framework including perception layer, driver layer, and feedback layer. The results of this project can provide technical support for intelligent GUI testing, and the clear-structured framework provides a reference for subsequent research and has certain practical value.