有没有一种创新的水印技术不会使水印变得明显？
有没有一种水印技术，既不使水印明显，又可以通过分析图像来识别来源？
Is there a watermarking technique which doesn't make the watermark obvious yet by analyzing the image you can identify the source?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（3）
是的，这被称为（图像的）“鲁棒”数字水印。即使在压缩、调整大小、裁剪或翻译图像时，强大的数字水印也应该能够被识别。因为您无法将强大的数字水印嵌入到最低阶位中（因为它们在压缩中丢失），所以该过程通常是
顺便说一句，为了使水印不明显但可识别，你只需要确保水印本身是随机的，这样你就无法立即识别它即使您执行了水印提取过程，如果您不知道水印应该是什么。
Yes, it's called "robust" digital watermarking (of images). A robust digital watermark should be recognizable even when you compress, resize, crop or translate the image. Because you can't embed robust digital watermarks into lowest-order bits, say (as they get lost in compression), the process is typically
BTW in order to make the watermark non-obvious but recognizable, you just need to make sure that the watermark in itself is random so that you can't immediately recognize it even if you carry out the watermark extraction procedure if you don't know that the watermark should be.
也许您可以使用隐写技术来隐藏图像低位中的签名？
（编辑：隐写术的缺点是图片的任何大小调整、重新编码或格式更改都可能会破坏信息。）
Perhaps you could employ steganographic tehniques to hide a signature in the lower-order bits of the image?
(Edit: A disadvantage of steganography is that any resizing, reencoding or format change of the picture is likely going to destroy the information.)
这是我的实现： https://www.facebook.com/ChromaShift/
它并不像与 Digimarc 条形码一样好，但运行良好，可以集成到 Web 应用程序中以动态“跟踪”JPEG。页面上有相当详细的白皮书和手册（pdf）文档。
This is my implementation: https://www.facebook.com/ChromaShift/
It is not as good as Digimarc Barcode but works well and can be integrated into web apps for dynamic "stegging" of JPEGs. There is a fairly detailed white paper and manual (pdf) document on the page.