Color recognition by extended color space method for 64-color 2-D barcode

Colorizing two-dimensional (2-D) barcodes is a promising modality for improvement on design and data capacity. The number of colors employed in recent 2-D color barcodes is limited less than 8. To increase the number of recognizable colors, we propose a color recognition method which classifies 64 colors, using an extended color space and pattern recognition techniques. The proposed method employs 4 seed colors embedded within 64-color code to prevent influences from environment illumination. In the proposed method, we extend the color representation for each code color in 2-D barcode from three components to higher number component. And a classifier is trained using the extended colors captured from several environmental lighting conditions. The classifier is selectable. In this paper, artificial neural network (ANN), support vector machine (SVM) and nearest neighbor (NN) are employed as the classifier. The experimental results shows that the proposed method achieves MSER less than 3.0%. Comparing the number of seed colors, 4 seed color is comparable to that using 8 seed color. SVM achieves the best recognition performance of 1.17% SER. These results suggest that 64-color recognition is possible by the proposed method.