摘要
Over exposure is rather annoying in photo taking. However, in some severe light conditions over exposure is inevitable using conventional cameras due to the limitation of dynamic range of the image sensor. The over exposed information would be completely lost and unrecoverable. In order to cope with this problem, we propose a novel technique in which the noise is used to enlarge the dynamic range of the image sensor. The essential mechanism that noise contributes to the information recovery is investigated. It is also proved that the visibility of regained information can reach the peak when specifically added noise is synchronized with the image sensor, thus activating the phenomenon of stochastic resonance （SR）. Four different types of noises are investigated to show the effects of variant distributions on the quality of recovered information. The experimental outcomes are consistent with our theoretical results, which indicates that the SR-based lost information recovery is quite promising.
Over exposure is rather annoying in photo taking. However, in some severe light conditions over exposure is inevitable using conventional cameras due to the limitation of dynamic range of the image sensor. The over exposed information would be completely lost and unrecoverable. In order to cope with this problem, we propose a novel technique in which the noise is used to enlarge the dynamic range of the image sensor. The essential mechanism that noise contributes to the information recovery is investigated. It is also proved that the visibility of regained information can reach the peak when specifically added noise is synchronized with the image sensor, thus activating the phenomenon of stochastic resonance （SR）. Four different types of noises are investigated to show the effects of variant distributions on the quality of recovered information. The experimental outcomes are consistent with our theoretical results, which indicates that the SR-based lost information recovery is quite promising.
基金
supported by the State Key Laboratory Fund of Fluid Power Transmission and Control
the National Natural Science Foundation of China (Grant Nos. 10932009 and 10972194)
