视错觉 画廊
探索令人叹为观止的视错觉——从旋转图案到尺寸扭曲。每种错觉都附有科学解释，揭示大脑被欺骗的原因。
Rotating Snakes
The brain interprets high-contrast luminance gradients as motion signals. The repeating dark-to-light pattern in adjacent rings triggers motion-detecting neurons in V5/MT cortex, creating the illusion of rotation even though nothing moves.
Müller-Lyer Illusion
The arrowheads create depth cues. Outward arrows suggest the line is at the far end of a corner (appearing shorter), while inward arrows suggest a near edge (appearing longer). The brain unconsciously scales perceived size to distance.
Ponzo Illusion
The converging lines mimic railroad tracks receding into the distance. Objects higher in the visual field appear further away, so the brain enlarges them to maintain size constancy — making the upper bar look larger.
Café Wall Illusion
The thin mortar lines between offset black and white tiles are perceived as tilted because adjacent tiles create local tilt signals in orientation-sensitive V1 neurons. The horizontal lines are perfectly parallel.
Hermann Grid
Ghost dots appear at intersections because lateral inhibition in retinal ganglion cells suppresses brightness signals at junctions. The surrounding dark squares reduce the inhibition effect at non-fixated points, creating ghost dark spots.
Checker Shadow
The brain interprets color relative to context, not in absolute terms. It assumes the shadow darkens square B, so it compensates and perceives B as lighter than A — even though both squares reflect the same amount of light.
Rubin's Vase
The brain struggles to simultaneously process foreground and background. Figure-ground segregation is a competitive process — areas that appear convex or smaller tend to become figure, causing the percept to flip between vase and faces.
Necker Cube
The flat 2D drawing contains no depth cues to determine which face is in front. The brain alternates between two equally valid 3D interpretations every few seconds, reflecting rivalry between competing neural representations.
Spinning Dancer
The silhouette provides no depth cues about which leg is in front. Your brain arbitrarily assigns depth, causing the perceived direction of rotation to flip. Focusing on the shadow or foot can help switch the direction.
Ames Room
The room is actually trapezoidal — the left corner is much further away. But the room's walls, floor, and ceiling are carefully designed to look rectangular when viewed through a peephole, making a distant person appear giant.
Fraser Spiral
The arcs are actually perfect concentric circles, but the twisted cord pattern (alternating black and white segments) creates local tilt signals that shift the perceived curvature. The brain smoothly integrates these local signals into a spiral.
Zöllner Illusion
Short diagonal hatch marks on alternating sides of each long line bias local orientation detectors in V1. Adjacent lines with opposing hatch directions appear to converge or diverge, though all four long lines are perfectly parallel.