PreSCC: Robust Screen-Camera Communication via Pre-Correction Against Perspective Distortion and Illumination Interference

Screen-camera communication (SCC) systems are widely used for Quick Response (QR) codes and other barcodes, and they are also basic techniques for visible light communication (VLC). However, the reliability of SCC systems is limited by perspective distortion and illumination interference. Existing techniques often rely on complex receiver-side post-processing to mitigate perspective distortion and employ fixed channel coding to suppress errors, which remain unreliable in dynamic environments. PreSCC introduces a sender-side pre-correction framework to address these limitations. It follows a three-stage pipeline of pose estimation, prediction, and pre-correction. First, the transmitter estimates the receiver pose and constructs an explicit mapping from screen pixels to camera pixels. Based on this mapping, the prediction algorithm identifies regions affected by perspective distortion and strong illumination interference. The pre-correction module then derives a robust decodability region, selects a large coding region via a maximum inscribed circle search, maps this region back to the screen, and embeds a pre-corrected code pattern within it. This sender-side precorrection pipeline distinguishes PreSCC from prior receiverside homography-based SCC designs, while remaining compatible with off-the-shelf decoders and running in real time on the transmitter CPU. We implement PreSCC as a mobile application and evaluate it through simulations and extensive experiments. The results show that PreSCC substantially enlarges the decodable pose region and achieves consistently higher decoding robustness than traditional SCC, even under joint perspective distortion and illumination interference.