Balancing Privacy and Credibility in High‐Definition Maps: A Zero‐Knowledge Watermarking Algorithm Based on Compressed Sensing
Mingwang Zhang, Liming Zhang, Tao Tan, Yang Zhao-jun, Wenqing Lyu · Transactions in GIS · 2025
ABSTRACT With the rapid advancement of autonomous driving, the privacy and credibility of high‐definition (HD) maps, which serve as an essential foundation for driving safety, are receiving increasing attention. Traditional ciphertext‐domain digital watermarking technology encounters high computational overhead and risks of privacy leakage, making it challenging to balance data security, privacy protection, and trustworthiness verification. Against this background, a zero‐knowledge watermark (ZKW) algorithm based on compressed sensing is proposed. First, the high‐precision map data in OpenDrive format is dynamically encrypted using DNA‐based techniques to enhance data security and privacy. Secondly, to ensure the credibility of data verification, a zero‐knowledge watermark is generated using compressed sensing and embedded into the attribute values of ciphertext‐domain data as invisible characters. Experimental results demonstrate that the proposed ZKW scheme is commutative with the encryption scheme and can achieve zero‐knowledge proof (ZKP) in both ciphertext and plaintext domains. Furthermore, the scheme exhibits excellent robustness against various security threats, including geometric attacks, cropping attacks, and combined attacks.