Color barcode decoding in the presence of specular reflection

Abstract

Color barcodes enable higher information density with respect to traditional black and white barcodes. Existing technologies use small color palettes and display the colors in the palette in the barcode itself for easy and robust decoding. This solution comes at the cost of reduced information density due to the fact that the displayed reference colors cannot be used to encode information. We introduce a new approach to color barcode decoding that uses a relatively large palettes (up to 24 colors) and a small number of reference colors (2 to 6) to be displayed in a barcode. Our decoding method specifically accounts for specular reflections using a dichromatic model. The experimental results show that our decoding algorithm achieves higher information rate with a very low probability of decoding error compared to previous approaches that use a color palette for decoding.