An image blur estimation scheme for severe out of focus images while scanning linear barcodes

In recent years, charge coupled device are used for scanning linear barcode. The main drawback of CCD is, it does not have the potential to handle the out-of-focus images. In this project, low resolution images with severe out-of-focus blur is resolved in spatial domain. This work is based on dynamic template matching scheme which make use of directed graphical model in spatial domain. The directed graphical model is used to determine the relationship between state variable from blurred waveform at specific blur level and observation sequence. A varying program based inference algorithm is used to recover the optimal state sequence and hence the system works in real-time.