Self-occlusion occurs when one part of an object or a character's body blocks the view of another part of themselves from the current camera angle or viewpoint. Because 3D objects are projected onto a 2D medium (such as a screen or canvas), geometry that lies further along the line of sight is naturally obscured by geometry that is closer to the viewer.

In 2D and 3D art, self-occlusion is a fundamental byproduct of perspective, depth, and pose. It is commonly used to construct dynamic poses, convey foreshortening, or tactfully obscure parts of a character's body.

Not to be confused with:

• AI generation artifacts or artistry errors - Instances where overlapping limbs or body parts unnaturally merge, clip through each other, or disappear entirely due to mistakes by the artist or generative algorithms, rather than being naturally or intentionally hidden by perspective.
• ambient_occlusion - A rendering method that calculates how exposed a point is to ambient light, generating soft shadows in crevices. While related to geometry, it is a lighting effect rather than the literal blocking of the viewer's line of sight.
• foreshortening - The visual contraction of an object or limb angled toward the viewer. Although foreshortening often results in self-occlusion, it refers to the scaling distortion rather than the literal blocking of geometry.
• out_of_frame - When parts of a character are hidden because they lie outside the boundaries of the image canvas.

See also:

• tag_group:perspective