Huang, Chun-HsiangChun-HsiangHuangJA-LING WU2018-09-102018-09-102009http://www.scopus.com/inward/record.url?eid=2-s2.0-58049110896&partnerID=MN8TOARShttp://scholars.lib.ntu.edu.tw/handle/123456789/348139Watermarking performance enhancement has always been a difficult task since the performance metrics of watermarking systems, i.e., fidelity, robustness, and payload size, inherently conflict with each other. Nowadays, most watermarking schemes hide payloads according to predefined rules or empirical perceptual models. Therefore, the performance of digital watermarking schemes can be determined only passively. In this study, a genetic algorithm-based framework for watermarking performance enhancement is proposed. Watermarked results having better robustness, guaranteed fidelity, and fixed payload size can be obtained. Existing blind-detection watermarking schemes can be improved significantly by incorporating the proposed framework. In addition, the proposed framework has many desirable advantages such as asymmetric embedding/detection overhead, easy integration with existing data-hiding schemes, and direct control over fidelity and robustness. © 2008 Elsevier Inc. All rights reserved.application/pdf1483892 bytesapplication/pdfDigital watermarking; Fidelity; Genetic algorithms (GAs); RobustnessControl theory; Diesel engines; Digital watermarking; Genetic algorithms; Watermarking; Direct controls; Fidelity; Payload sizes; Perceptual models; Performance enhancements; Performance metrics; Robustness; Robustness enhancements; Watermarking schemes; Watermarking systems; Papermakingjournal article10.1016/j.ins.2008.10.0352-s2.0-58049110896