Bit error prediction for digital image data

A nonideal two-dimensional optical system, as encountered in digital holographic data storage applications, can modify the intensity of transmitted digital data through beam shaping, focal surface distortion, and moire patterns. Such changes in intensity can have significant adverse effects on digital data recovery at the receiver (e.g., a CCD camera). Current research seeks to detect and correct classes of such distortion so that recovery methods can be applied to the received data. This paper discusses methods used to predict the locations of bit errors in the recovered data. Prediction information may be used as weighting information in the recovery algorithm and in the design of channel codes. Furthermore, the higher the level of distortion that can be tolerated in the system, the lower the cost of the corresponding lenses, making the system more tractable for commercialization.