Description
In laser line scan (LLS) imaging systems, frame and line synchronization between source and receiver is critical for image formation. However, in a truly multi-static LLS system, the illuminator and the receiver are physically separated at different locations/platforms, which introduces the risk of source and receivers losing sync with each other. On the other hand, in the image data, adjacent lines/pixels in general are highly correlated, and so re-establishing the sync using image data content is feasible. The raw data recorded at the receiver of a LLS imager is a periodic data stream.
Line sync detection relies on determining the fundamental period of the data stream. This is similar to pitch detection in speech signal processing. Therefore, the average magnitude difference function (AMDF) pitch detector is a suitable candidate to solve this problem:
N is the window size used in computation and K1 to Km are the range of pitches to be examined. Initially, AMDF Minimum is the fundamental pitch. Since there is no multiplication required, it is hardware-implementation friendly.
Featureless sand patches were imaged to validate the effectiveness of this algorithm:
( a) Featureless sand patch images (b) 1D data stream of the sand patch
For more information, please contact Bing Ouyang
