shapefromX

Shape from X refers to a family of computer vision techniques used to infer the three-dimensional shape of an object or scene from one or more two-dimensional images. The "X" in Shape from X represents the type of visual information used as input. Common examples include Shape from Shading, Shape from Texture, Shape from Defocus, and Shape from Silhouette.

Shape from Shading attempts to recover shape by analyzing how light interacts with surfaces. It assumes a model of how light reflects off a surface and uses the intensity variations in an image to infer the surface normal at each point, which can then be integrated to reconstruct the 3D shape. This method is sensitive to lighting conditions and surface properties.

Shape from Texture utilizes the perceived distortion of a repeating texture pattern in an image to infer surface orientation and curvature. When a textured surface is viewed from different angles or has varying curvature, the texture elements appear stretched, compressed, or skewed, providing cues about the underlying geometry.

Shape from Defocus leverages the fact that objects at different distances from a camera will have different degrees of blur. By analyzing the amount of defocus in an image, or by using images taken with different focal settings, it's possible to estimate the depth of points in the scene and reconstruct its 3D shape.

Shape from Silhouette, also known as Shape from Outline, reconstructs 3D shape from the 2D outline or silhouette of an object. If multiple silhouettes are available from different viewpoints, the intersection of the corresponding 3D cones can be used to generate a volumetric approximation of the object's shape. These methods are often employed in 3D scanning and object reconstruction applications.