Spatial resolution of 1D image-based barcode reading : an information theoretical approach

We evaluate in this paper the minimal bar width required to recover the binary information of a barcode captured by a linear camera. After estimating the average mutual information of a 1D image-based system we show that this barcode resolution can easily be derived from the signal to noise ratio. By placing the study within the field of information theory, average mutual information defines a novel performance criterion which supplements classical optical criteria as it links the minimal barcode density to the amount of effectively recoverable bits.