Deep Convolutional Neural Network (CNN) is trained by large-scale labelled datasets. After training, the model can achieve high recognition rate or good classification effect. However, the training of CNN models with smaller-scale datasets usually occurs overfitting. In order to solve this problem, a novel data augmentation method called GMM-CGAN was proposed, which was integrated Gaussian Mixture Model (GMM) and CGAN (Conditional Generative Adversarial Net). Firstly, sample number was increased by randomly sliding sampling around the core region. Secondly, the random noise vector was supposed to submit to the distribution of GMM model, then it was used as the initial input to the CGAN generator and the image label was used as the CGAN condition to train the parameters of the CGAN and GMM models. Finally, the trained CGAN was used to generate a new dataset that matched the real distribution of the samples. The dataset was divided into 12 classes of 386 items. After implementing GMM-CGAN on the dataset, the total number of the new dataset was 38600. The experimental results show that compared with CNN's training datasets augmented by Affine transformation or CGAN, the average classification accuracy of the proposed method is 89.1%, which is improved by 18.2% and 14.1%, respectively.

To solve the problems of "center-surround" salient object detection and extraction method, such as incomplete object detected or extracted, not smooth boundary and redundancy caused by down-sampling 9-level pyramid, a salient object detection method based on Reciprocal Function and Spectral Residual (RFSR) was proposed. Firstly, the difference between the intensity image and its corresponding Gaussian low-pass one was used to substitute the normalization of the intensity image under "center-surround" model, meanwhile the level of Gaussian pyramid was further reduced to 6 to avoid redundancy. Secondly, a reciprocal function filter was used to extract local orientation information instead of Gabor filter. Thirdly, spectral residual algorithm was used to extract spectral feature. Finally, three extracted features were properly combined to generate the final saliency map. The experimental results on two mostly common benchmark datasets show that compared with "center-surround" and spectral residual models, the proposed method significantly improves the precision, recall and F-measure, furthermore lays a foundation for subsequent image analysis, object recognition, visual-attention-based image retrieval and so on.