SCIA2009_superresolution - Spatio-temporal Super-Resolution...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
Spatio-temporal Super-Resolution Using Depth Map Yusaku Awatsu, Norihiko Kawai, Tomokazu Sato, and Naokazu Yokoya Graduate School of Information Science, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma, Nara 630-0192, Japan Abstract. This paper describes a spatio-temporal super-resolution method using depth maps for static scenes. In the proposed method, the depth maps are used as the parameters to determine the corresponding pixels in multiple input images by assuming that intrinsic and extrinsic camera parameters are known. Because the proposed method can deter- mine the corresponding pixels in multiple images by a one-dimensional search for the depth values without the planar assumption that is of- ten used in the literature, spatial resolution can be increased even for complex scenes. In addition, since we can use multiple frames, temporal resolution can be increased even when large parts of the image are oc- cluded in the adjacent frame. In experiments, the validity of the proposed method is demonstrated by generating spatio-temporal super-resolution images for both synthetic and real movies. Keywords: Super-resolution, Depth map, View interpolation. 1 Introduction A technology that enables users to virtually experience a remote site is called telepresence [1]. In a telepresence system, it is important to provide users with high spatial and high temporal resolution images in order to make users feel like they are existing at the remote site. Therefore, many methods that increase spatial and temporal resolution have been proposed. The methods that increase spatial resolution can be generally classiFed into methods that use one image as input [2,3] and methods that require multiple images as input [4,5,6,7]. The methods using one image are further classiFed into two types: ones that need a database [2] and ones that do not [3]. The former method increases the spatial resolution of the low resolution image based on previous learning of the correlation between various pairs of low and high resolution images. The latter method increases the spatial resolution by using a local statistic. These methods are e±ective for limited scenes but largely depend on the database and the scene. The methods using multiple images increase the spatial resolution by corresponding pixels in the multiple images that are taken from di±erent positions. These methods determine pixel values in the super- resolved image by blending the corresponding pixel values [4,5,6] or minimizing A.-B. Salberg, J.Y. Hardeberg, and R. Jenssen (Eds.): SCIA 2009, LNCS 5575, pp. 696–705, 2009. c ± Springer-Verlag Berlin Heidelberg 2009
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Spatio-temporal Super-Resolution Using Depth Map 697 the diference between the pixel values in an input image and the low resolution image generated From the estimated super-resolved image [7]. Both methods require the correspondence oF pixels with sub-pixel accuracy. However, in these methods, the target scene is quite limited because the constraints oF objects in
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 04/22/2010 for the course MI IP taught by Professor Vladbalan during the Spring '10 term at Universidad del Rosario.

Page1 / 10

SCIA2009_superresolution - Spatio-temporal Super-Resolution...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online