为什么云端背景移除存在隐私风险
remove.bg、Canva、Adobe Express 这类云端抠图工具会把你的照片上传到它们的服务器，照片在那里可能被存储、记录，或用于训练它们的模型——对于证件照、家庭照或任何含有可识别人脸的图片来说，这都是不必要的暴露。基于浏览器的工具则能在图片从不离开设备的情况下抠除背景。（关于 AI 分割的实际工作原理以及更全面的隐私对比，详见 AI 背景去除的 Learn 文章。）
五步去除背景
- 1打开 ClearCut，通过拖入或点击浏览来添加你的图片——它支持 JPEG、PNG 和 WebP。首次使用时，AI 分割模型会下载到你的浏览器中（一次性的短暂加载）；之后它便能即时、离线运行。
- 2让 AI 抠出主体。它会分析每一个像素并自动去除背景——对于主体清晰、背景分明的图片，一两秒就能得到干净的结果。
- 3查看左右并排的预览，并用手动画笔修补任何遗漏之处——把自动处理丢失的细发丝、眼镜或纤细边缘重新涂回来，或擦除残留的背景碎片。
- 4可选地把抠出的图放到新背景上：保持透明、填入纯色（电商标准是白色），或合成到另一张图片上。
- 5把结果下载为用于设计工具的透明 PNG，或下载为更小巧、适合网页的 WebP。一切都留在你浏览器的内存中，并在你关闭标签页的那一刻清除。
获得更佳效果的技巧
As a general rule, photos with strong contrast between the subject and the background produce the best cutout results. A person wearing a dark jacket against a light-colored wall is an ideal example; a person dressed in camouflage standing in dense foliage is far more challenging to separate cleanly. For product photography, it is strongly recommended that you shoot against a plain white or solid-color background — this lets the AI segment the subject from the background especially cleanly, with near-perfect edges. If the automatic detection accidentally misses parts of your subject (which is especially common with hair, animal fur, or semi-transparent objects like glass), you can simply use the manual brush tool to paint those missed regions back in, stroke by stroke. One more thing worth remembering: higher-resolution source images generally produce cleaner edges, and this is particularly noticeable for extremely fine details such as individual strands of hair.