: In our daily life, people are frequently confused by the problem of not understanding the target information related to the professional field. To address the problem, a natural scene information acquisition and analysis model based on deep learning is proposed in this study, which is divided into four parts. In turn, Connectionist Text Proposal Network text detection algorithm, projection-based text segmentation, Capsule network text recognition, text analysis methods are adopted to process the target information, so that users can get intelligent solutions targeted at professional problems through uploading target information pictures. In this paper, the model is employed to analyze the nutritional intake of different users to achieve precision medicine. Moreover, the proposed model can accurately identify the nutrient component table, find the relationship between nutrients and diseases, and provide dietary advice for users.

For tunnel deformation analysis using traditional measurement methods to obtain tunnel section data, there are problems such as small data coverage and low efficiency. 3D laser scanning technology has the advantages of automatic, high precision and high efficiency in collecting the point cloud data of the target object, and can completely and accurately express the target entity. Based on the tunnel point cloud acquired by 3D laser scanner, the tunnel engineering modeling research is carried out in this paper. Firstly, the engineering survey and 3D reconstruction technology of shield tunnel were carried out based on 3D laser scanning technology. The total station layout control was used to scan the subway platform and tunnel interval, and the high-precision laser point cloud data were obtained. Secondly, a random sampling consistency algorithm is proposed to extract engineering measurement results such as tunnel axis and cross section. Finally, the 3D modeling of the tunnel is established by using the stretching setting out modeling method. Taking Yuzhu Tunnel planning and acceptance project as an example, the experimental results show that the method can effectively visualize the overall deformation of the tunnel, and provide accurate and scientific spatial data for tunnel engineering measurement, operation and maintenance.

At present, there are a large of container ports in the Greater Bay Area,which is existing some problems,such as the overlapping functionality of the port,lack of scientific positioning,fierce competition,lack of coordination mechanism and so on.In order to accelerate the integration between the ports,form completely advantages,win-win cooperation,production factors configuration of optimal new development mode,to make the balance between the port. This paper will based on the current situation of the Greater Bay Area to pursue the research about the theory of systematic cluster analysis and the advantages and disadvantages in its application area. The data sources of the container ports in the Greater Bay Area will be used for the experiment and eleven ports’ classification and stratification with SPSS software to systematic analysis. In the end,this paper will combine with the experience of coordinated development of key port groups at home and abroad to provide feasible,reliable and credible solutions and suggestions for the construction of the port group in the Guangdong-Hong Kong-Macao Greater Bay Area.This study has theoretical guidance and practical significance for the coordinated development of container ports in the Guangdong-Hong Kong-Macao Greater Bay Area.

There are problems such as low recognition accuracy and large classification error in the existing classification methods for ship identification based on optical remote sensing images. In this paper, we will analyze the characteristics of ships and determine the indicative factors for applying remote sensing to monitor ships in combination with optical remote sensing images. Using optical remote sensing image data, combined with U-Net and AttU-Net deep neural network models, we assist in extracting new remote sensing indices with strong generality and clear physical meaning, and establishing rules for monitoring ships, so as to establish a more general and clear physical meaning of the monitoring and identification method of remote sensing satellite images. The method is applied and evaluated with port optical remote sensing image data. The data show that compared with traditional machine learning methods, the accuracy of ship monitoring using U-Net and AttU-Net deep learning models in this paper reaches 89.04%, and the recall rate and accuracy rate are better than SVM. it shows that the model can detect ships effectively.