Calibration of DIDSON forward-scan acoustic video camera

Computer image processing to extract scene and target information is critical in automated underwater search and inspection, target detection, tracking, mapping and object recognition. In turbid waters, the new generation of acoustic video cameras, e.g., a DIDSON, can produce images with higher target details than available from optical sensors. Calibration of these cameras to establish agreement between the data and image projection model is the prerequisite for the accurate estimation of quantitative information, including positioning and image registration. This paper addresses the internal calibration problem, in order to estimate and rectify the lens distortion in a DIDSON camera to ensure data to model consistency