Acoustic Camera Viewer

How a microphone-array beamforming sound map overlays onto an image. Pick a scene, drag Heatmap opacity to fade the map over the photo.

Workflow & algorithm → the full pipeline: sensors → beamforming → road-plane geometry → lane attribution → ANPR fusion Config / calibration → each camera's 4-point ground homography (world ↔ pixel) + SoundCam pose

Heatmap opacity 70%

low → high sound level (relative dB)

Explain — how the raw data maps to the image

1 · The microphone array
where the 56 mics physically sit (metres)

raw 56-channel waveform — **2 · The raw data**
56 channels × samples — pressure vs time, one row per mic

3 · Delay-and-sum (the mapping) — click the image above to focus the array there. Each mic is shifted by its travel-time delay, then all are summed.

focus: — m

delay spread across array: —

beam power — (0 dB = on the source)

Raw — as recorded (arrivals misaligned)

Aligned — shifted by delay (coherent when on source)

Beamformer output = sum of the aligned channels — tall when the mics agree, flat when they cancel