Reproduction of sound signal from gramophone records using three-dimensional scene reconstruction

Preserving the invaluable historic recording on gramophone records, whether it is speech or music, makes a lot of sense in our societies, because the traditional record play back system wears out the record gradually or even destroys some of the most fragile sample in the archives.;An optical flow based framework for a record playing system is presented. Two classic optical flow algorithms by Lucas and Kanade and Horn and Schunck were implemented and compared. A modification to the Horn and Schunck's optical flow algorithm was made using a groove orientation spatial constraint on the record groove and its optical flow distribution characteristics and was used in this work.;Four recent 3D reconstruction algorithms (that recover relative depth/surface orientation from optical flow/image intensity gradients) were investigated and compared quantitatively through synthesised scenes and the best algorithm (by Barron, Ngai and Spies) was used in the work presented in this thesis.;A robust estimation method of computing the surface orientation was built into the framework following the 3D reconstruction process.;This thesis presents a non-contact method to reproduce sound signal from gramophone records using 3D scene reconstruction of the micro-grooves carved on a record surface.;The microscope that is used for imaging was modified to overcome the limitation of a shallow depth of field by using a thin glass plate to obtain part of the image at a second focal length to gain better overall quality images of the groove.;The sound signal was recovered from the surface orientation and processed further using the industry standard RIAA (Recording Industry Artists Association) filter which was also implemented as a digital filter.;The overall algorithm has been tested and found to be working correctly using both simulated data of record groove images and that of real groove images.