KEYWORDS: RGB color model, Image restoration, Convolution, Data modeling, Feature selection, Signal to noise ratio, Process modeling, Network architectures
This paper considers how to restore an image with large stained areas. An end to end model is proposed, which contains two networks, an edge generation adversarial network and a content generative adversarial network. First, the edge generation adversarial network is deployed to infer the missing boundaries of the image. Then the second network with a designed edge information channel is employed to restore the missing or stained areas of the image with the guidance of the inferred boundaries. Experiments were performed on ImageNet. The results show that the proposed model can better understand the semantic information of the stained area by introducing additional object contour channels and greatly improve the inpainting capability of the model. Quantitative evaluation indexes show that the proposed model is 4.5% better than the DeepFill V2 model in structural similarity and 7.1% better than the DeepFill V2 model in Peak Signal-to-Noise Ratio.
Cloud detection is important for the application of space-borne video remote sensing. Video data of Chinese Jilin-1 is detected through migration learning and improved Unet with fully connected conditional random field. Due to the interference of fast movement of cloud targets and satellite platform jitter in video satellite remote sensing, it is difficult for Unet network depth to effectively extract the context characteristics of cloud targets, and effect of segmentation and cloud detection is poor. To solve the problem of missing cloud target extraction features, this paper uses the VGG16 pretraining model as the backbone network of the context path, and refines the segmentation results using the fully connected conditional random field (Fully Connected / Dense CRF) to improve cloud boundary pixel localization. The results show that the proposed algorithm can effectively improve the model segmentation accuracy, where the accuracy and intersection ratio reach 92.6% and 90.9%. The proposed network has strong generalization and high practical application value.
The aim of style transfer is giving the style from one picture to another. The application of neural network in image processing separates the high level features and low level features of the image in the process of style transfer, and derives a variety of methods and optimization for style processing. The style transfer generates new images by separating and recombining the content and style of original images. In this process, various factors such as color and illumination will affect the result. The traditional algorithm only focuses on continuous pixels and the whole image, this paper will extend the process object to the contour of the image, and improves the detail processing from the existing style transfer examples. From the contour of images, the target image retains the contour feature of style image and the content of original image, in other word, gives the contour style of style image to original image. Finally, the style transfer effect based on the original image contour is obtained with some defects. The work can be easily extended to the aspects of video and 3D images.
Since the labels of training samples are related to bags not instances, the multiple instance learning (MIL) is a special ambiguous learning paradigm. In this paper, we propose a novel bag space (BS) construction and extreme learning machine (ELM) combination method named BS_ELM for MIL, which can capture the bag structure and use the efficiency of ELM. Firstly, sparse subspace clustering is performed to obtain the cluster centers and a new bag space is constructed. Then ELM is used to classify bags in the new space. Experiments on data sets demonstrate the utility and efficiency of the proposed approach as compared to the other state-of-the-art MIL algorithms.
With the development of optical engineering technology, the precision of 3D scanning equipment becomes higher, and its role in 3D modeling is getting more distinctive. This paper proposed a 3D scanning modeling method that has been successfully applied in Chinese ancient city reconstruction. On one hand, for the existing architectures, an improved algorithm based on multiple scanning is adopted. Firstly, two pieces of scanning data were rough rigid registered using spherical displacers and vertex clustering method. Secondly, a global weighted ICP (iterative closest points) method is used to achieve a fine rigid registration. On the other hand, for the buildings which have already disappeared, an exemplar-driven algorithm for rapid modeling was proposed. Based on the 3D scanning technology and the historical data, a system approach was proposed for 3D modeling and virtual display of ancient city.
KEYWORDS: 3D modeling, Cultural heritage, Data modeling, 3D metrology, Distance measurement, Calibration, Clouds, Cameras, 3D image processing, Reconstruction algorithms
Cultural Heritage reflects the human production, life style and environmental conditions of various historical periods. It exists as one of the major national carriers of national history and culture. In order to do better protection and utilization for these cultural heritages, a system of three-dimensional (3D) reconstruction and statistical measurement is proposed in this paper. The system solves the problems of cultural heritage’s data storage, measurement and analysis. Firstly, for the high precision modeling and measurement problems, range data registration and integration algorithm used to achieve high precision 3D reconstruction. Secondly, multi-view stereo reconstruction method is used to solve the problem of rapid reconstruction by procedures such as the original image data pre-processing, camera calibration, point cloud modeling. At last, the artifacts’ measure underlying database is established by calculating the measurements of the 3D model’s surface. These measurements contain Euclidean distance between the points on the surface, geodesic distance between the points, normal and curvature in each point, superficial area of a region, volume of model’s part and some other measurements. These measurements provide a basis for carrying out information mining of cultural heritage. The system has been applied to the applications of 3D modeling, data measurement of the Terracotta Warriors relics, Tibetan architecture and some other relics.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.