EM-Based joint symbol and blur estimation for 2D barcode

Noura Dridi (1) , Yves Delignon (1) , Wadih Sawaya (1) , Christelle Garnier (1)

1 LAGIS-SI

Decoding a severely blurred 2D barcode can be considered as a special case of blind image restoration issue. In this paper, we propose an appropriate system model which includes the original image with the particularities related to barcode, the blur and the observed image. We develop an unsupervised algorithm that jointly estimates the blur and detects the symbols using the maximum likelihood (ML) criterion. Besides, we show that when taking into account the spatial properties of the barcode, the prohibitive complexity of the ML algorithm can be reduced without degrading its performance. Simulation results show that the algorithm performs accurate estimation of the blur and achieves good performance for symbol detection which is close to that obtained with supervised algorithm.

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Noura Dridi, Yves Delignon, Wadih Sawaya, Christelle Garnier. EM-Based joint symbol and blur estimation for 2D barcode. International Symposium on Image and Signal Processing and Analysis (ISPA), Sep 2011, France. pp.32-36. ⟨hal-00805753⟩