Zhenlong Man, Xinyu Zhou, Shuping Li
Abstract In the contemporary digital era, the unauthorised disclosure and manipulation of fingerprint images belonging to individuals with disabilities can result in irreparable consequences, including identity theft and significant privacy infringements. Of particular concern is the potential misuse of their unique pathological areas as identity markers, which can render this group vulnerable to discriminatory data mining and tracking practices. In order to address the concerns related to information security that pertain to incomplete fingerprints of individuals with disabilities, this paper puts forth a protection scheme that integrates anomaly filling and robust encryption. In order to ensure the security of pathological regions, the proposed scheme identifies anomalous areas through the use of local standard deviation and employs an enhanced 2D-ECAM chaotic system to generate random noise for obfuscation filling at the damaged sites. Furthermore, it presents an efficient randomization algorithm that simulates molecular irregular motion and designs a novel chain diffusion strategy incorporating the domino effect, thereby achieving comprehensive protection for incomplete fingerprint images. Experimental results demonstrate that this solution effectively conceals abnormal fingerprint information. The ciphertext image achieves an information entropy of 7.9983, with number of pixel change ratio and uniform average change intensity both reaching theoretical optimal values. These results fully validate the algorithm’s outstanding resistance to statistical and differential attacks, providing effective security and indistinguishability for fingerprint images of individuals with disabilities.