GUI-Based Automated Testing Tool with Exploratory Behavioral Analysis for Enhanced Software Quality Assurance

Sasikala Devireddy,S M K Sukumar Reddy,K.G.S. Venkatesan
DOI: https://doi.org/10.58599/ijsmien.2023.1902
2023-09-30
Abstract:Testing web-based apps is difficult because of the intricate nature of these programmes. Automated methods used in computerised testing make ensuring that tasks are distributed fairly and reduce the influence of human mistake. This article’s goal is to provide our system for automating the testing of web applications. It was possible to effectively construct this state-of-the-art automated testing framework thanks to the Selenium WebDriver tool. As a result of this setup, PyAutoGUI may be used to automatically traverse through websites and match widgets to a pre-trained dataset. With the framework’s screenshot feature, designers may examine their previous work with more ease. Our team has come up with a novel approach to finding, categorising, and rating widgets by using machine learning and image processing techniques. Using this method, you may quickly get the images needed for the GUI widgets from the server by inputting the URL for the training data. Datasets are checked for the presence of GUI widgets using BLOB text detection. After locating all widgets, they are classified according to their various uses. Widgets may take several forms, including labels, buttons, input fields, checkboxes, and links. Finally, we make sure the overall GUI component categorization is as precise as possible by performing a series of checks. In order to identify the different kinds of widgets, our project used image processing rather than graphical user interfaces coded in languages like Java, Python, or Visual Basic. Test coverage was increased to 98.4 percent after we successfully implemented automated testing for online apps.
What problem does this paper attempt to address?