Liver is the main organ of human metabolic function. At present, the main problems of machine learning in the semantic segmentation of liver images are as follows:1) there are inferior vena cava, soft tissue and blood vessels in the middle of the liver, and even some necrosis or hepatic fissures; 2) the boundary between the liver and some adjacent organs is blurred and difficult to distinguish. In order to solve the problems mentioned above, the Edge Attention Model (EAM)and the Edge Attention Net (EANet) were proposed by using Encoder-Decoder framework. In the encoder, the residual network ResNet34 pre-trained on ImageNet and the EAM were utilized, so as to fully obtain the detailed feature information of liver edge; in the decoder, the deconvolution operation and the proposed EAM were used to perform the feature extraction to the useful information, thereby obtaining the semantic segmentation diagram of liver image. Finally, the smoothing was performed to the segmentation images with a lot of noise. Comparison experiments with AHCNet were conducted on three datasets, and the results showed that:on 3Dircadb dataset, the Volumetric Overlap Error (VOE) and Relative Volume Difference (RVD) of EANet were decreased by 1.95 percentage points and 0.11 percentage points respectively, and the DICE accuracy was increased by 1.58 percentage points; on Sliver07 dataset, the VOE, Maximum Surface Distance (MSD) and Root Mean Square Surface Distance (RMSD) of EANet were decreased approximately by 1 percentage points, 3.3 mm and 0.2 mm respectively; on clinical MRI liver image dataset of a hospital, the VOE and RVD of EANet were decreased by 0.88 percentage points and 0.31 percentage points respectively, and the DICE accuracy was increased by 1.48 percentage points. Experimental results indicate that the proposed EANet has good segmentation effect of liver image.

Concerning the problem that the local feature and rotation invariant feature cannot be jointly paid attention to in traditional attention-based neural networks, a multi-branch neural network model based weakly supervised fine-grained image classification method was proposed. Firstly, the lightweight Class Activation Map (CAM) network was utilized to localize the local region with potential semantic information, and the residual network ResNet-50 with deformable convolution and Oriented Response Network (ORN) with rotation invariant coding were designed. Secondly, the pre-trained model was employed to initialize the feature networks respectively, and the original image and the above regions were input to fine-tune the model. Finally, the three intra-branch losses and between-branch losses were combined to optimize the entire network, and the classification and prediction were performed on the test set. The proposed method achieves the classification accuracies of 87.7% and 90.8% on CUB-200-2011 dataset and FGVC_Aircraft dataset respectively, which are increased by 1.2 percentage points, and 0.9 percentage points respectively compared with those of the Multi-Attention Convolutional Neural Network (MA-CNN) method. On Aircraft_2 dataset, the proposed method reaches 91.8% classification accuracy, which is 4.1 percentage points higher than that of ResNet-50. The experimental results show that the proposed method improves the accuracy of weakly supervised fine-grained image classification effectively.

Although the existing image inpainting methods can recover the content of the missing area of the image, there are still some problems, such as structure distortion, texture blurring and content discontinuity, so that the inpainted images cannot meet people’s visual requirements. To solve these problems, an image inpainting method based on dilated convolution was proposed. By introducing the idea of dilated convolution to increase the receptive field, the quality of image inpainting was improved. This method was based on the idea of Generative Adversarial Network (GAN), which was divided into generative network and adversarial network. The generative network included global content inpainting network and local detail inpainting network, and gated convolution was used to realize the dynamical learning of the image features, solving the problem that the traditional convolution neural network method was not able to complete the large irregular missing areas well. Firstly, the global content inpainting network was used to obtain an initial content completion result, and then the local texture details were repaired by the local detail inpainting network. The adversarial network was composed of SN-PatchGAN discriminator, and was used to evaluate the image inpainting effect. Experimental results show that compared with the current image inpainting methods, the proposed method has great improvement in Peak Signal-to-Noise Ratio (PSNR), Structural SIMilarity (SSIM) and inception score. Moreover, the method effectively solves the problem of texture blurring in traditional inpainting methods, and meets people’s visual requirements better, verifying the validity and feasibility of the proposed method.

Current recognition algorithm of pneumonia image faces two problems. First, the extracted features can not fit the pneumonia image well because the transfer learning model used by the pneumonia feature extractor has large image difference between the source dataset and the pneumonia dataset. Second, the softmax classifier used by the algorithm can not well process high-dimensional features, and there is still room for improvement in recognition accuracy. Aiming at these two problems, a pneumonia image recognition algorithm based on Deep Convolution Neural Network (DCNN) was proposed. Firstly, the GoogLeNet Inception V3 network model trained by ImageNet dataset was used to extract the features. Then, a feature fusion layer was added and random forest classifier was used to classify and forecast. Experiments were implemented on Chest X-Ray Images pneumonia standard dataset. The experimental results show that the recognition accuracy, sensitivity and specificity of the proposed model reach 96.77%, 97.56% and 94.26% respectively. The proposed model is 1.26 percentage points and 1.46 percentage points higher than the classic GoogLeNet Inception V3+Data Augmentation (GIV+DA) algorithm in the index of recognition accuracy and sensitivity, and is close to the optimal result of GIV+DA in the index of specificity.

In order to improve the learning accuracy of Auto-Encoder (AE) algorithm and further reduce the classification error rate, Sparse marginalized Denoising Auto-Encoder (SmDAE) was proposed combined with Sparse Auto-Encoder (SAE) and marginalized Denoising Auto-Encoder (mDAE). SmDAE is an auto-encoder which was added the constraint conditions of SAE and mDAE and has the characteristics of SAE and mDAE, so as to enhance the ability of deep learning. Experimental results show that SmDAE outperforms both SAE and mDAE in the given classification tasks; comparative experiments with Convolutional Neural Network (CNN) show that SmDAE with marginalized denoising and a more robust model outperforms convolutional neural network.

This paper proposed a monitoring and online quality analysis method based on real-time data analysis in order to solve the problems that data stream is hard to manage and analyze and production monitoring and online quality analysis can not be handled effectively in the process of steel heating furnace production. By combining real-time database and relational database, and using six Sigma management tools and control chart techniques, the authors proposed an approach to do monitoring and online quality analysis. The implemented system includes real-time data processing, production monitoring as well as online/off-line quality monitoring. The performance of the system indicates that it can be effectively applied in the heating furnace in real-time data analysis and online quality monitoring.