Subpixel rendering

About: Subpixel rendering is a research topic. Over the lifetime, 3885 publications have been published within this topic receiving 82789 citations.

A Learning Based Depth Estimation Framework for 4D Densely and Sparsely Sampled Light Fields

Abstract: This paper proposes a learning based solution to disparity (depth) estimation for either densely or sparsely sampled light fields. Disparity between stereo pairs among a sparse subset of anchor views is first estimated by a fine-tuned FlowNet 2.0 network adapted to disparity prediction task. These coarse estimates are fused by exploiting the photo-consistency warping error, and refined by a Multi-view Stereo Refinement Network (MSRNet). The propagation of disparity from anchor viewpoints towards other viewpoints is performed by an occlusion-aware soft 3D reconstruction method. The experiments show that, both for dense and sparse light fields, our algorithm outperforms significantly the state-of-the-art algorithms, especially for subpixel accuracy.

High‐resolution digital elevation models derived from Viking Orbiter images: Method and comparison with Mars Orbiter Laser Altimeter Data

Abstract: The purpose of this paper is to present a method to derive local and high-resolution digital elevation models (DEM) of Mars from Viking Orbiter images. We focus on two aspects that appear to be new in photogrammetry applied to Mars: the correlation method and the relative orientation method based on the analysis of the coplaneity. We demonstrate that a DEM from two Viking Orbiter images selected according to the criteria of parallax height ratio, resolution, and illumination conditions can be obtained with a better spatial lateral resolution than the interpolated Mars Orbiter Laser Altimeter (MOLA) data. Two images on the Deuteronilus Mensae area satisfy these criteria. The algorithm of reconstruction of the topography is divided into two steps. First, a correlation method at subpixel level based on Fourier transform has been adapted to extract image location of homologous points. As initial camera positions and angles are known with poor accuracy, an adjustment is then carried out by the method of relative orientation which requires no absolute control points and takes advantage of the subpixel accuracy of our matching algorithm. Finally, the DEM is projected into a Martian reference frame and registered to MOLA data for validation. The spatial resolution depends on the roughness of the area and is locally close to the distance between individual MOLA foot prints along one profile (300 m), as demonstrated from the comparison with MOLA profiles. These first results of high-resolution DEM from Viking Orbiter images should be used with MOLA data for many geological studies that require high-resolution topographic data.

Supbixel rendering suitable for updating an image with a new portion

Abstract: In an image update, a display apparatus receives only a new portion ( 1110 ) of an image for display but does not receive the remaining, unchanged portion of the image. The display apparatus performs a subpixel rendering (SPR) operation ( 454 ) for the new portion but does not redo the SPR for the whole image. Efficient techniques are provided to achieve good appearance at the edges between the new portion and the rest of the image. Other features are also provided.

Discrimination and identification for subpixel targets in hyperspectral imagery

Abstract: Spectral measures have been used in material identification and discrimination. They are effective if the spectral signatures are calibrated and not contaminated. However, it may not be true in many real applications, specifically, for mixed pixels and subpixel targets. This paper investigates the issue of discrimination and identification for subpixel targets and further develops sample spectral covariance/correlation matrix-based hyperspectral measures to account for spectral variability within subpixel targets. Two types of measures are of interest and studied, Mahalanobis distance-based hyperspectral measures and matched filter-based hyperspectral measures. In order to substantiate the proposed measures, a real data-based comparative analysis is conducted and compared to two spectral similarity measures, spectral angle mapper (SAM) and spectral information divergence (SID) for performance evaluation. The experiments show that both Mahalanobis distance-based hyperspectral measures and matched filter-based hyperspectral measures work very effectively and outperformed the SAM and the SID in discrimination and identification for subpixel targets.