作者: James Z. Wang , Ritendra Datta , Jia Li
DOI:
关键词:
摘要: In a system and method for the generation of attack-resistant, user-friendly, image-based CAPTCHAs (Completely Automated Public test to Tell Computers Humans Apart), controlled distortions are applied randomly chosen images presented user annotation from given list words. An image is that contains multiple connected but independent with borders between them distorted or otherwise visually obfuscated in way computer cannot distinguish selects near center one images. The performed satisfies incongruous requirements low perceptual degradation high resistance attack by content-based retrieval systems. Word choices carefully generated avoid ambiguity as well attacks based on themselves.