Abstract
We present a novel joint barcode deblurring and nonuniform illumination compensation algorithm for barcode signals whose number of modules is known. The proposed algorithm is based on a penalized least squares method using a roughness penalty function for an illumination model and a double well penalty function for a barcode signal model. In simulations, the proposed method shows an improved performance compared with a conventional method without compensating nonuniform illumination effects. In addition, the proposed method converges quickly during optimization (within 15 iterations), thereby showing strong possibility for real time decoding of barcode signals.
| Original language | English |
|---|---|
| Pages (from-to) | 718-724 |
| Number of pages | 7 |
| Journal | Transactions of the Korean Institute of Electrical Engineers |
| Volume | 57 |
| Issue number | 4 |
| State | Published - Apr 2008 |
Keywords
- Barcode
- Deblurring
- Illumination
- Module
- Regularization