Поиск ошибок в программах для обработки графических изображений с использованием метода фаззинга

Abstract

Increasing number of software for automated graphics processing requires effective testing methods. One of these methods is fuzzing, for which it is necessary to determine the most effective algorithms for creating test data in order to increase the number of errors found and minimize hardware resources. The comparison of algorithms for creating test data for finding errors in the executable code of programs designed for processing graphic images is the result of the performed research. Using Bayesian networks to describe fuzzing allows determining the relationships between structural components during testing. Based on the results of the comparison of fuzzing algorithms for creating test data, the most effective algorithms for finding errors in the executable code of programs for processing graphic images have been identified. The performance of the proposed algorithms was tested on a number of existing vulnerabilities classified as CVE (Common Vulnerabilities and Exposures). The processing of the results of experiments on the creation of test data was carried out using the simulation environment, allowing analyzing the testing process step by step. The obtained research results, algorithms for creating test data for finding errors can be used at various stages of software testing