知识卡片

普通读书笔记卡 · 9013 · 生成模型:GPT-6

图像表示决定显式结构

理解等级:中级

规则像素数组把位置藏在索引中,便于卷积;像素集合 {(ℓ_i,x_i,y_i)} 显式携带位置,便于几何变换和不规则采样;坐标函数 ℓ(x,y)=f_θ(x,y) 则通过参数表达连续位置的颜色。三者可配合使用,但转换回显示网格时需要采样与插值。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      However, an image can be represented in different ways, making explicit certain aspects of the information present on the input. What else could we do to represent an image? Another very different, but equivalent, representation is to encode an image as a collection of points, indicating its color and location explicitly. In this case, an image is represented as a set of pixels:

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第38章《图像与几何表示》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c038b.xhtml) - 结论依据:摘录中关于图像表示决定显式结构的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:However, an image can be represented in different ways, making explicit certain aspects of the information present on the input. What else could we do to represent an image? Another very different, but equivalent, representation is to encode an image as a collection of points, indicating its color and location explicitly. In this case, an image is represented as a set of pixels: