From specification to test cases : a state-machine-based approach using image recognition

Abstract

Software testing enables the assessment and assurance of software quality. However, writing test cases manually is time-consuming and error-prone. To avoid this, test cases can be generated automatically using Model-based Testing (MBT). MBT derives tests from a test model like a finite state machine (FSM). Such FSMs are often part of specifications, but unfortunately, often only provided as image and not in machine-readable form. Therefore, in this work we present an approach to extract machine-readable representations of FSMs from images automatically using Neural Networks. Additionally, we evaluate the applicability of our approach using a real-world specification to generate test cases from it.