Test data generation for the image processing domain

Abstract

In the context of software development, validation, verification and in particular testing activities, have received much attention from developers and, in consequence, researchers seeking to define techniques and tools that allow such activities to be carried out systematically, with high quality and low cost. In this scenario, the automation of testing activities is crucial, since its manual execution is unproductive and subject to errors. Part of the automation process is related to the generation of test data, i.e., the choice of input data that meet requirements defined by a particular test technique. The automation of data generation is not a trivial task and many studies have been published on the subject. Even more complicated becomes this problem when the domain is complex. This proposal seeks to collaborate on this theme, making use of optimization techniquesbased on metaheuristic search to generate test data in the context of programs for image processing. Thus, it fits in the context of software engineering but carries important elements of the image processing and computational intelligence areas. (AU)