← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8897.b · 生成模型:GPT-6

视觉滤波模型保留直流

理解等级:中级

本章用 h=−Δg+λg 粗略模拟视觉的空间频率响应:负 LoG 提供中心周边作用,λg 保留低频;若 g 归一化,连续模型的直流增益为 λ。其与对比敏感函数的相似性是定性近似,不能视为完整神经模型或个体视觉测量。

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

      The Laplacian filter can also be used as a coarse approximation of the behavior of the early visual system. When looking at the magnitude of the DFT of the Laplacian with σ = 1 (figure 18.17), the shape seems reminiscent of our subjective evaluation of our own visual sensitivity to spatial frequencies when we look at the Campbell and Robson chart (figure 16.21). As the

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第18章《图像导数》(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/c018.xhtml) - 结论依据:摘录中关于视觉滤波模型保留直流的定义、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:The Laplacian filter can also be used as a coarse approximation of the behavior of the early visual system. When looking at the magnitude of the DFT of the Laplacian with σ = 1 (figure 18.17), the shape seems reminiscent of our subjective evaluation of our own visual sensitivity to spatial frequencies when we look at the Campbell and Robson chart (figure 16.21). As the