In zero-visibility water — turbid harbors, silted construction sites, flooded tunnels — optical cameras see nothing. Imaging sonar sends acoustic pulses and constructs images from the reflections. But a single sonar produces a flat 2-D image. To map in three dimensions, you need multiple sensors, expensive phased arrays, or time-consuming multi-pass surveys.
A single imaging sonar can produce 2.5-D acoustic maps by reading the shadows.
When a sonar pulse illuminates an object, the object casts an acoustic shadow on the far side — a dark region where no reflected signal returns. The geometry of that shadow encodes the object's height. A taller object casts a longer shadow. The relationship is trigonometric: shadow length, sonar grazing angle, and object height form a triangle that can be solved from the image alone.
The method treats acoustic shadows not as noise to be filtered out but as information — specifically, the missing signal carries the dimensional data that the reflected signal cannot. Presence tells you where things are. Absence tells you how tall they are.
Pool experiments demonstrated reliable spatial layout and dimension reconstruction from a single, cheap sensor. The practical applications are immediate: port security inspection, underwater construction monitoring, disaster response in flooded structures, archaeological surveys in murky water.
The through-claim: in any imaging system, the dark regions of the image contain information about what cast them. The instinct to maximize signal everywhere — to illuminate, to enhance, to fill in the gaps — can destroy information that only exists in the absence of signal. Sometimes the shadow is the measurement, and the light is just the context that makes the shadow interpretable.