Aiming at the problem of model information leakage caused by interpretability in Deep Neural Network （DNN）， the feasibility of using the Gradient-weighted Class Activation Mapping （Grad-CAM） interpretation method to generate adversarial samples in a white-box environment was proved， moreover， an untargeted black-box attack algorithm named dynamic genetic algorithm was proposed. In the algorithm， first， the fitness function was improved according to the changing relationship between the interpretation area and the positions of the disturbed pixels. Then， through multiple rounds of genetic algorithm， the disturbance value was continuously reduced while increasing the number of the disturbed pixels， and the set of result coordinates of each round would be maintained and used in the next round of iteration until the perturbed pixel set caused the predicted label to be flipped without exceeding the perturbation boundary. In the experiment part， the average attack success rate under the AlexNet， VGG-19， ResNet-50 and SqueezeNet models of the proposed algorithm was 92.88%， which was increased by 16.53 percentage points compared with that of One pixel algorithm， although with the running time increased by 8% compared with that of One pixel algorithm. In addition， in a shorter running time， the proposed algorithm had the success rate higher than the Adaptive Fast Gradient Sign Method （Ada-FGSM） algorithm by 3.18 percentage points， higher than the Projection & Probability-driven Black-box Attack （PPBA） algorithm by 8.63 percentage points， and not much different from Boundary-attack algorithm. The results show that the dynamic genetic algorithm based on the interpretation method can effectively execute the adversarial attack.

In order to retrieve images of traffic violations from a large number of road traffic images， a semantic retrieval method based on specific object self-recognition was proposed. Firstly， road traffic domain ontology as well as road traffic rule description were established by experts in traffic domain. Secondly， traffic image features were extracted by Convolutional Neural Network （CNN）， and combined with the strategy for classifying image features which is based on the proposed improved Support Vector Machine based Decision Tree （SVM-DT） algorithm， the specific objects and the spatial positional relationship between the objects in the traffic images were automatically recognized and mapped into the association relationship （rule instance） between the corresponding ontology instance and its objects. Finally， the image semantic retrieval result was obtained by reasoning based on ontology instances and rule instances. Experimental results show that the proposed method has higher accuracy， recall and retrieval efficiency compared to keyword and ontology traffic image semantic retrieval methods.

In response to the issue that data anomalies in the uncertain databases often hamper the efficient and effective use of data, an uncertain data cleaning method was proposed to reduce abnormal data based on the information entropy theory. First, the uncertainty degree of uncertain data was defined by using information entropy. Then, the confidence interval of uncertain data was obtained based on statistical method with the degree of uncertainty. By means of the confidence interval, the uncertain databases were cleaned. The experimental results show the effectiveness and efficiency of the proposed method.

In order to correct the barrel distortion of wide-angle lens, a new method for distortion correction for barrel distortion was proposed. Adopting the lattice model calibration method, according to the location relation of dots between distortion image and ideal figure, the offset surfaces in X-and Y-axis direction about distorted pixels were got. Then, the cubic B-spline interpolation function was adopted to interpolate the surface. Thus, the offsets of each pixel were obtained in the distorted image. Furthermore, the pixels shift was rectified to achieve an undistorted image by coordinate conversion. And then the bilinear interpolation was used to reconstruct the gray level of pixels. The simulation results show that the proposed method can make a good correction of the coordinate position and gray value.