Directional Moving Averaging Interpolation for Texture Mapping
✍ Scribed by Chung-Lin Huang; Kou-Chang Chen
- Publisher
- Elsevier Science
- Year
- 1996
- Tongue
- English
- Weight
- 990 KB
- Volume
- 58
- Category
- Article
- ISSN
- 1077-3169
No coin nor oath required. For personal study only.
✦ Synopsis
The cheapest texture filtering method is point sampling, where the pixel nearest the desired sampled point is used.
In texture mapping, several filtering techniques have been proposed which allow prefiltering of a texture. It has been However, for stretched images, the texture pixels are visidemonstrated that the shape-variant filter kernel has better ble as large blocks, causing jagged edges, and for shrunken performance than its shape-invariant counterpart. However, images, aliasing can cause distracting moire patterns. These the conventional shape-variant filtering in which the filtering phenomena are called ''aliasing. '' Tang and Suen [4] have area is warping-based uses the same kernel for high-detail areas proposed five particular image transformation models and as for low-detail areas of the image in texture space. In this two general transform models. However, the transformed paper we propose a hybrid interpolation filtering method for unfiltered image has jagged edges problems. Aliasing retexture mapping which is based on content-based space-variant sults when a signal has unreproducible high frequencies.
directional filtering concepts. First, we classify the local image
To avoid aliasing, we must low-pass filter the input signal area in the texture space into three types of blocks: constant, to make it band-limited before sampling. However, since oriented, and irregular blocks. Second, for constant and irregular blocks, we use local average filtering and elliptical weighted the input signal is discrete, we must reconstruct the continaverage (EWA) filtering, respectively. Third, for oriented uous signal from input samples by convolution.
blocks, we propose a new filtering method called elliptical
Various algorithms [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] for image interpolation have weighted adaptive directional moving average (EWADMA) been proposed. Among them, the fast B-spline transform filtering. In the experiments, we show that our method algorithm using the recursive moving average proposed by may recover the distorted images with better subjective Unser et al. [6] has two major advantages. First, it asymptotquality.