傅里叶变换和最大值
有没有一种方法可以在知道两个函数 (f,g) 的傅里叶变换的情况下有效地计算它们的最大值的傅里叶变换?
Is there a way to compute efficiently the Fourier transform of the max of two functions (f,g), knowing their Fourier transform?
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)
我对此表示怀疑。当且仅当 |f| 的傅里叶变换可以有效地计算 max(f, g) 的傅里叶变换可以有效地计算。 (因为 max(f,g) = (f+g+|fg|)/2。)
但F{f}和F{|f|}之间似乎没有任何关系...
I doubt it. The Fourier transform of max(f, g) can be computed efficiently if and only if the Fourier transform of |f| can be computed efficiently. (Because max(f,g) = (f+g+|f-g|)/2.)
But there seems to be no relationship between F{f} and F{|f|}...
假设您指的是每个点的最大值,并且由于最大值是非线性运算,因此没有任何方法可以做到这一点。您需要在时域中进行最大值运算,然后执行傅立叶变换。
Assuming you mean the max at each point, and since max is a non-linear operation, there is not going to be any way to do this. You would need to do the max operation in the time domain and then perform the Fourier transform.
您可以使用 FT(max(f(x),g(x))) = FT( H(fg)f + (1-H(fg))f) = FT( H(fg) f) + FT( 1-H(fg))f)
但在这里你会陷入困境,因为据我所知,不存在如此酷的两个函数组合公式。
即使你知道
FT(f)
FT(g)
FT(fg)=FT(f)-FT(g)
FT(H)=1/2(delta+1/(pi is))
只需打开积分并算出如何组合术语。它可以是令人难以置信的困难,也可以是令人难以置信的简单。试试吧。
You can stat with FT(max(f(x),g(x))) = FT( H(f-g)f + (1-H(f-g))f) = FT( H(f-g) f) + FT(1-H(f-g))f)
But here you will stucked because as I know there are no exist such cool formulas for compoistion of two functions.
Even if you know
FT(f)
FT(g)
FT(f-g)=FT(f)-FT(g)
FT(H)=1/2(delta+1/(pi i s))
Just open integral and figure out how to combine terms. It can be unbelievalbe hard or unbelievale simple. Just try.