Access the full text.
Sign up today, get DeepDyve free for 14 days.
K Jia, X Wang, X Tang (2013)
Image transformation based on learning dictionaries across image spacesIEEE Trans. Pattern Anal. Mach. Intell., 35
Z Cui, H Chang, S Shan, B Zhong, X Chen (2014)
European Conference on Computer Vision
R Hartley, A Zisserman (2003)
Multiple View Geometry in Computer Vision
N Snavely, SM Seitz, R Szeliski (2006)
Photo tourism: exploring photo collections in 3DACM Trans. Graph. (TOG), 25
G Freedman, R Fattal (2011)
Image and video upscaling from local self-examplesACM Trans. Graph. (TOG), 30
J Yang, Z Wang, Z Lin, S Cohen, T Huang (2012)
Coupled dictionary training for image super-resolutionIEEE Trans. Image Process., 21
C Dong, CC Loy, K He, X Tang (2015)
Image super-resolution using deep convolutional networksIEEE Trans. Pattern Anal. Mach. Intell., 38
CR Qi, H Su, K Mo, LJ Guibas (2017)
Deep learning on point sets for 3D classification and segmentationIEEE Proc. Comput. Vis. Pattern Recogn. (CVPR), 1
Y Liu, Y Xiong (2008)
Automatic segmentation of unorganized noisy point clouds based on the gaussian mapComput. Aided Des., 40
L Kobbelt, M Botsch (2004)
A survey of point-based techniques in computer graphicsComput. Graph., 28
J Yang, J Wright, TS Huang, Y Ma (2010)
Image super-resolution via sparse representationIEEE Trans. Image Process., 19
D Nistér (2004)
An efficient solution to the five-point relative pose problemIEEE Trans. Pattern Anal. Mach. Intell., 26
KI Kim, Y Kwon (2010)
Single-image super-resolution using sparse regression and natural image priorIEEE Trans. Pattern Anal. Mach. Intell., 32
DG Lowe (2004)
Distinctive image features from scale-invariant keypointsInt. J. Comput. Vis., 60
Current state-of-the-art point rendering techniques such as splat rendering generally require very high-resolution point clouds in order to create high-quality photo realistic renderings. These can be very time consuming to acquire and oftentimes also require high-end expensive scanners. This paper proposes a novel deep learning-based approach that can generate high-resolution photo realistic point renderings from low-resolution point clouds. More specifically, we propose to use co-registered high-quality photographs as the ground truth data to train the deep neural network for point-based rendering. The proposed method can generate high-quality point rendering images very efficiently and can be used for interactive navigation of large-scale 3D scenes as well as image-based localization. Extensive quantitative evaluations on both synthetic and real datasets show that the proposed method outperforms state-of-the-art methods.
The Visual Computer – Springer Journals
Published: May 11, 2018
Access the full text.
Sign up today, get DeepDyve free for 14 days.