A study on depth up-sampling and refinement

Autor: Chien-Ting Li, 李乾鼎
Rok vydání: 2011
Druh dokumentu: 學位論文 ; thesis
Popis: 99
In this thesis, we develop a method to interpolate the depth and then refine the depth sequence based on maximum apriori probability. The proposed method is composed of two parts: pre-processing and depth refinement. In the pre processing, we align the color images and the depth sequence. In the depth refinement, we enlarge the size of the depth sequence and then refine the depth values. To refine the depth value, we simultaneously consider the color, spatial, and temporal information by using maximum apriori probability. Experimental results show that the proposed depth refinement method can function well compared with traditional methods such as low-pass filtering and median filtering.
Databáze: Networked Digital Library of Theses & Dissertations