| 摘要 |
PURPOSE: A system and a method for rendering three-dimensional views of a scene are provided to preserve a plurality of soft edges of objects within rendered three-dimensional views. CONSTITUTION: A system configures an alpha model of one or more randomly selected areas related to one or more objects within a depth map(104). The system presumes a color value, a depth value, and a transparent value about pixels of identified mixed pixel areas(112-116). The system generates three-dimensional representation of a scene based on multilayered points(118). The system renders one or more three-dimensional views of the scene(120). [Reference numerals] (102) Processing a depth map of a scene; (104) Configuring an alpha model of a randomly selected area related to one or more objects having a depth value similar to the depth map and the randomly selected area; (106) Identifying important edges related to an object inside the depth map of a scene; (108) Identifying pixels related to the identified edges; (110) Identifying a foreground area, a background area, and mixed pixel area related to the identified edges; (112) Estimating a depth value about each pixel of the identified mixed pixel area; (114) Estimating a color value about each pixel of the identified mixed pixel area; (116) Estimating a transparent value about each pixel of the identified mixed pixel area; (118) Generating a three-dimensional expression of a scene based on a multi-layer point; (120) Rendering three-dimensional views of a scene by using the three-dimensional expression based on the multi-layer point |
