Automated visual testing of HTML pages

Abstract

Web applications displayed on heterogeneous platforms and have a frequently changing life cycle. There is a high demand for their visual testing. This paper presents an approach for automated visual regression testing. The basic concepts of visual testing are discussed. The traditional, manual testing checks the displayed content of various browsers and platforms. This is a huge amount of monotonous work, which often results in human mistakes thus reducing reliability and consistency. Another disadvantage of the manual testing is that it is slow. Certain web applications have weekly updates, manual testing is not a feasible option for them. Automation would gain us consistency and be free from human errors. Obviously, the configuration of the test environment takes some effort but that could return on long term. The paper investigates testing time and reliability of automated visual regression testing in a selected environment.