Colour brings extra data capacity with regards to QR codes, however, it likewise conveys tremendous challenges to the decoding because of color interface and illumination variation, particularly for high-density QR codes. In this project, we put forth a system for high-capacity QR codes, HiQ, which optimizes the decoding algorithm for high-density QR codes to accomplish robust and quick decoding on cell phones, and receives a learning-based approach for color recovery. In addition, we propose a robust geometric transformation algorithm to adjust the geometric distortion. We likewise give a challenging color QR code dataset, CUHK-CQRC, which comprises of 5390 high-density color QR code tests caught by various cell phones under various lighting conditions. Experimental results demonstrate that HiQ out performance the benchmark by 286% in decoding success rate and 60% in bit error rate.