Deep Neural Network (DNN) is proved to be vulnerable to adversarial sample attacks in many key deep learning systems such as face recognition and intelligent driving. And the detection of various types of adversarial samples has problems of insufficient detection and low detection efficiency. Therefore, a deep learning model oriented adversarial sample difference detection method was proposed. Firstly, the residual neural network model commonly used in industrial production was constructed as the model of the adversarial sample generation and detection system. Then, multiple kinds of adversarial attacks were used to attack the deep learning model to generate adversarial sample groups. Finally, a sample difference detection system was constructed, containing total 7 adversarial sample difference detection methods in sample confidence detection, perception detection and anti-interference degree detection. Empirical research was carried out by the constructed method on the MNIST and Cifar-10 datasets. The results show that the adversarial samples belonging to different adversarial attacks have obvious differences in the performance detection on confidence, perception and anti-interference degrees, for example, in the detection of confidence and anti-interference, the adversarial samples with excellent performance indicators in perception show significant insufficiencies compared to other types of adversarial samples. At the same time, it is proved that there is consistency of the differences in the two datasets. By using this detection method, the comprehensiveness and diversity of the model's detection of adversarial samples can be effectively improved.

To reduce the navigation graph size and redundant test paths in the generation process of Web application test cases, a Web application test case generation method based on Selenium page object design pattern and graph traversal algorithm was proposed. Firstly, by classifying the original page objects, the page object navigation graph with navigation page object class as the node and the navigation method as the migration edge was created. Secondly, with the shortest-path algorithm of graph, a Page Object Graph Algorithm (POGA) was proposed to realize the navigation graph traversal in order to generate test path set. Finally, the test paths were extracted and Faker was used to generate the simulated data, and the test cases that can be directly executed were produced. The experimental results show that, the proposed method has the reduction rate of about 89% compared with the navigation graph size generated by crawling Web applications, reduces the number of redundant and infeasible paths in comparison with the state migration method for generating Web application test cases, and further improves the reuse rate of page objects and the maintainability of test cases.

Aiming at the low efficiency of test case automatic generation technology, an IMproved Bacterial Foraging Optimization Algorithm (IM-BFOA) was proposed with introduction of Knet map. Firstly, Kent map was used to increase the diversity of the initial population and global search of bacteria. Secondly, the step size of chemotaxis stage in the algorithm was adaptively designed to make it more rational in the process of bacterial chemotaxis. Finally, a fitness function was constructed according to the program under test to accelerate the optimization of test data. The experimental results show that compared with Genetic Algorithm (GA), Particle Swarm Optimization (PSO) algorithm and standard Bacterial Foraging Optimization Algorithm (BFOA), the proposed algorithm is the best in terms of iterations number and running time with the guarantee of coverage and has high efficiency of test case generation.

In view of the problem that the test cost is increased by a large number of test suites in regression testing, a test suite reduction method based on weak mutation criterion was proposed. Firstly, the relation matrix between test suites and mutation branches was obtained based on weak mutation criterion. Then, four invalid test requirements and subset test suites were reduced repeatedly. Finally, the current optimal test suite was selected by using artificial fish swarm algorithm, and the simplification and test suite selection operations were performed alternately until all the test requirements were covered. Compared with Greedy algorithm and HGS (Harrold-Gupta-Soff) algorithm on six classical programs, when using weak mutation criterion with no changing or slightly changing mutation score, the reduction rate was improved by 73.4% and 8.2% respectively, and the time consumption was decreased by 25.3% and 56.1% respectively. The experimental results show that the proposed method can effectively reduce the test suites and save the test cost in regression testing.

Concerning the low accuracy of existing software birthmark extraction algorithms in detecting code theft problem, a new static software birthmark extraction algorithm was proposed. The birthmark generated by this algorithm covered two kinds of software features. The source program and the suspicious program were preprocessed to get the program meta data, which was used to generate Application Programming Interface (API) call set and instruction sequence as two features. These two features were synthesized to generate software birthmarks. Finally, the similarity of source program and suspicious program was calculated to determine whether there was code theft between the two programs. The experimental result verifies that the birthmark combined by API call set and instruction sequence has credibility and resilience, and has stronger resilience compared with k-gram birthmark.

The scale of test suites is constantly expanding and the cost of testing is increasing due to the change of test requirements in the process of regression testing. In order to solve the problems, a Reduction method of Test suites based on the analysis of Mutation (RTM) was proposed. Firstly, the test suites were classified and the transaction set matrix of mutants was created in binary numerical form according to whether the designated mutants could be detected or not by test suites. Then, the correlation relation between test suites was obtained by using the improved association mining algorithm. Finally, the test suites were effectively reduced according to these relations. The simulation experimental results of the six classical programs show that, the test suite reduction rate of the proposed RTM can reach 37%. Compared with the traditional greedy algorithm and heuristic algorithm, the proposed RTM improves the test suite reduction rate by 6%, and can guarantee the test coverage rate at the same time, even the test coverage rate of a single test suite increases by 11% on average. The proposed method can meet more test requirements by using fewer test suites, effectively improving test efficiency and reducing test cost.

Focusing on the false positive problem of null pointer dereference (NPD) defect in static testing, a new false positive recognition method for null pointer reference defect based on classification was proposed. The knowledge of NPD defect was mined and preprocessed to generate data set of the defects. Then the data set of NPD defects was classified via ID3 classification algorithm based on rough set theory, and there were two kinds of classification results, one was false positive null pointer reference defect instances, the other was real null pointer reference defect instances. The real NPD defects were confirmed according to the classification results of the defect instances by recognizing the false positive NPD defects. The method was tested on ten benchmark programs and compared to the NPD defect detection method based on the mainstream static testing tool FindBugs, the false positive rate was reduced by 25%, and the confirmation amount was reduced by 24% for NPD defects. The experimental result shows that the proposed method can effectively reduce defect confirmation overhead and improve the detection efficiency and stability for NPD defects in static testing.

It is difficult for the original test data to meet the requirements of the new version of software testing in regression testing, thus a new test data augmentation method based on Adaptive Particle Swarm Optimization (APSO) algorithm was proposed to solve the problem. Firstly, according to the similarity between the cross path and the target path of the original test data in the new version of the program, the appropriate test data in the original test data was chosen as evolutionary individual of initial population. Secondly, taking advantage of different sub-paths of the cross path of initial test data and target path, the input component which caused deviation between them was confirmed. Finally, the fitness function was created according to the path similarity, and the new data was generated by using the APSO algorithm to operate the input component. Compared with the genetic algorithm based and random algorithm based test data augmentation methods on four benchmark programs, the augmentation efficiency of the proposed method was improved on average by approximately 56% and 81% respectively. The experimental results show that the proposed method can effectively increase the efficiency and improve the stability of test data augmentation in regression testing.

Aiming at the error accumulation problem in the process of the traditional global matching of the three-dimensional (3D) models, a global optimal matching method based on swarm intelligences was proposed. The global matching process for multiple 3D fragments was abstracted, and then a mathematic model of the global optimal matching was set up, the solution of the optimal matching for multiple 3D fragments was converted to satisfy certain constraint conditions of the optimal match matrix of combinatorial optimization problem. A discretization algorithm based on hybrid social cognitive optimization algorithm was proposed to solve the NP (Non-deterministic Polynomial) problem. Finally, the classical example analyses verified that the proposed algorithm has global optimization ability and strong robustness without the initial position, and it provides an efficient method for global matching of the 3D fragments.

Aiming at the problem of erroneous judgement resulted from the lack of permission and time sequence considerations in traditional Web test case generation, a method combining the Unified Modeling Language (UML) activity diagram and statechart diagram was proposed,which generated testing cases according to the different users' permissions and interaction process analysis of Web page links. This proposed method generated extended statechart diagram containing information elements, got the final Web link testing cases with consideration of permission and sequence through transforming the extended statechart diagram and reordering the corresponding paths. In the comparison with the traditional Web testing case generation method which lacked the permission and time sequence considerations, this method avoids the erroneous judgement effectively and has obvious advantages in coverage, accuracy and efficiency.The experimental results show that the proposed method can improve the efficiency, reliability and feasibility of the Web test.

Due to low error checking rate of Web application test, a method of test case generation for Web applications based on state transition was proposed. By constructing state transition diagram of pages, event transition table and navigation transition table, the link relationship of Web applications was shown. This approach generated test path from state transition tree of pages got from state transition diagram of pages. Based on equivalence partitioning principles, a coverage criteria was proposed, then a test case set was reported as result combined with information from event transition table and navigation transition table. The result shows that the proposed method can represent link relationship of Web applications effectively, and improve error checking rate of test case.