The main reasons that can decrease the quality of recognition:
- low resolution or too small barcode (not enough pixels represent the barcode)
- noise (the barcode's fields are covered by irregularly placed points in background color)
- geometrical distortions (e.g. low resolution re-sampling at scanning, faxing)
- lossy image compression method applied
- color depth decreased with dithering algorithm.
It is straightforward, that the recognition's precision increases, if
- enough pixels represent the parts of barcodes with properly chosen resolution.
- decrease the noise e.g. with proper illumination
- good quality, high resolution imaging devices
- only lossless image compression method is applied (avoid lossy JPEG)
- do not allow any kind of dithering method at decreasing the color depth.