Improving contrast accuracy and resolution of laser speckle contrast imaging using two-dimensional entropy algorithm

A two-dimensional entropy laser speckle contrast imaging (E-LSCI) algorithm is proposed to improve the accuracy and resolution of laser speckle contrast imaging (LSCI) systems. Owing to the stable characteristics of two-dimensional entropy in speckle images, it can be used as a reference to correct contrast. E-LSCI introduces two-dimensional entropy into the contrast ratio calculation. It uses this benchmark and performs normalization to obtain a more accurate contrast ratio. The results of the method validation experiment show that the accuracy and contrast resolution of E-LSCI are higher than those of traditional LSCI by 27% and over 80%, respectively. The results of a finger blocking experiment show that E-LSCI can reduce the variance of contrast ratio at an identical flow velocity and improve the resolution by 25%. The application of this algorithm can improve the contrast accuracy and contrast resolution of laser speckle contrast imaging instruments.