Fiber Optic specklegram sensor based on image processing using LoG filter

Abstract

In fiber optic sensors, laser source produces a coherent light which is transmitted through a fiber cable. The output of this laser forms a speckle. This speckle is captured by a CCD and then analyzed by a PC. The intensity and shape of the speckle change when external force is applied on the cable. In this paper, Laplacian of Gaussian (LoG) filter is used to improve the dynamic range and accuracy of the fiber optic sensor. Capturing images is done in a dark room and the simulation is done by Matlab. The difference between the intensities of a normal and abnormal state image was 4.0556 when using 2D Discrete Wavelet transform (DWT) filter. It is increased into 14.6191 when LoG filter is used, thus increasing the dynamic range and accuracy of the sensor. In addition, the time is decreased by about 24.5 %. The number of changed blocks is increased by about 22 % leading to double increase in the accuracy and dynamic range.