deconvolution programm
- To: mathgroup at smc.vnet.net
- Subject: [mg4074] deconvolution programm
- From: nik at ksc.iasnet.ru
- Date: Fri, 31 May 1996 01:34:15 -0400
- Sender: owner-wri-mathgroup at wolfram.com
I would like an evaluation of the following notebook.
Nikolai Berdunov
e-mail: nik at ksc.iasnet.ru
>Do you want this posted to the Mathematica newsgroup and mailing
>list?
>Steve Christensen
> (*^
>
> ::[ frontEndVersion = "Microsoft Windows Mathematica Notebook Front End Version 2.2";
> microsoftWindowsStandardFontEncoding;
> fontset = title, "Arial", 24, L0, center, nohscroll, bold;
> fontset = subtitle, "Arial", 18, L0, center, nohscroll, bold;
> fontset = subsubtitle, "Arial", 14, L0, center, nohscroll, bold;
> fontset = section, "Arial", 14, L0, bold, grayBox;
> fontset = subsection, "Arial", 12, L0, bold, blackBox;
> fontset = subsubsection, "Arial", 10, L0, bold, whiteBox;
> fontset = text, "Arial", 12, L0;
> fontset = smalltext, "Arial", 10, L0;
> fontset = input, "Courier New", 12, L0, nowordwrap, bold;
> fontset = output, "Courier New", 12, L0, nowordwrap;
> fontset = message, "Courier New", 10, L0, nowordwrap, R65280;
> fontset = print, "Courier New", 10, L0, nowordwrap;
> fontset = info, "Courier New", 10, L0, nowordwrap;
> fontset = postscript, "Courier New", 8, L0, nowordwrap;
> fontset = name, "Arial", 10, L0, nohscroll, italic, B65280;
> fontset = header, "Times New Roman", 10, L0, right, nohscroll, italic;
> fontset = footer, "Times New Roman", 10, L0, right, nohscroll, italic;
> fontset = help, "Arial", 10, L0, nohscroll;
> fontset = clipboard, "Arial", 12, L0, nohscroll;
> fontset = completions, "Arial", 12, L0, nowordwrap, nohscroll;
> fontset = graphics, "Courier New", 10, L0, nowordwrap, nohscroll;
> fontset = special1, "Arial", 12, L0, nowordwrap, nohscroll;
> fontset = special2, "Arial", 12, L0, center, nowordwrap, nohscroll;
> fontset = special3, "Arial", 12, L0, right, nowordwrap, nohscroll;
> fontset = special4, "Arial", 12, L0, nowordwrap, nohscroll;
> fontset = special5, "Arial", 12, L0, nowordwrap, nohscroll;
> fontset = leftheader, "Arial", 12, L0, nowordwrap, nohscroll;
> fontset = leftfooter, "Arial", 12, L0, nowordwrap, nohscroll;
> fontset = reserved1, "Courier New", 10, L0, nowordwrap, nohscroll;]
> :[font = subtitle; inactive; nohscroll; center; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280;
fo
> ntColorRed = 0; fontColorGreen = 0; fontColorBlue = 0; bold; fontName = "Arial"; fontSize = 18; ]
> Three-Dimensional Deconvolution Programm for Reconstraction of Atomic Force Microscopy Image
>
> :[font = section; inactive; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280; fontColorRed = 0; fo
nt
> ColorGreen = 0; fontColorBlue = 0; bold; fontName = "Arial"; fontSize = 14; ]
> Discussion and Implementation
>
> :[font = text; inactive; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280; fontColorRed = 0; fontC
ol
> orGreen = 0; fontColorBlue = 0; plain; fontName = "Arial"; fontSize = 12; ]
> Well-knowing problem of Atomic Force Microscopy (AFM) technique is obtaining of the true surface image since AFM imag
e
> is a surface-tip convolution.
> Evidently that the greater is the Tip, the greater is Image distortion and conversely. And the less is structure wic
h
> we want observing, the less must be the Tip,It is hopeless case. We suggest the our algorithm until the problem is te
ch
> nology insoluble.
>
> :[font = subsection; inactive; ]
> Example
>
> :[font = text; inactive; startGroup; Cclosed; ]
> As an illustration our algorithm we use cubic structure on the plane
>
> :[font = input; endGroup; nowordwrap; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280; fontColorR
ed
> = 0; fontColorGreen = 0; fontColorBlue = 0; bold; fontName = "Courier New"; fontSize = 12; ]
> Nt=5; Ny=40; Nx=40;
> Tip=Table[(51.+(x^2+y^2)),{x,-Nt,Nt},{y,-Nt,Nt}];
> Samp=Table[If[i>15 && j>15 && i<25 && j<25,25,0],{i,Ny},{j,Nx}];
> New=Table[0,{i,Ny-2 Nt},{j,Nx-2 Nt}];
> l2:=Table[GrayLevel[1],{i,Ny-2 Nt-1},{j,Nx-2 Nt-1}];
> q1=ListPlot3D[Samp[[Range[Nt+1,Ny-Nt],Range[Nt+1,Nx-Nt]]],
> l2,PlotRange->{All,All,{-5,40}}];
> :[font = text; inactive; startGroup; Cclosed; ]
>
> and semispherical Tip.
> :[font = input; endGroup; nowordwrap; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280; fontColorR
ed
> = 0; fontColorGreen = 0; fontColorBlue = 0; bold; fontName = "Courier New"; fontSize = 12; ]
> Tip1=Table[(51.+(x^2+y^2)),{x,-2Nt,2Nt},{y,-2Nt,2Nt}];
> l2:=Table[GrayLevel[1],{i,4 Nt},{j,4 Nt}];
> q1=ListPlot3D[Tip1,l2,PlotRange->{All,All,{Min[Tip],140}},
> ViewPoint->{1.566,-2.890,0.803}, Boxed->False,Axes->False];
> :[font = text; inactive; startGroup; Cclosed; ]
> Then result of Tip-scaning on Surface.(It will be dependent on the cub and sphere size).
>
> :[font = input; endGroup; nowordwrap; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280; fontColorR
ed
> = 0; fontColorGreen = 0; fontColorBlue = 0; bold; fontName = "Courier New"; fontSize = 12; ]
> Do[New[[i,j]]=Min[N[Tip-Samp[[Range[i,2 Nt+i],
> Range[j,2Nt+j]]]]],{i,Ny-2 Nt},{j,Nx-2 Nt}];
> l2:=Table[GrayLevel[1],{i,Ny-2 Nt-1},{j,Nx-2 Nt-1}];
> q1=ListPlot3D[Max[New]-New,l2,PlotRange->{All,All,{-5,40}}];
> :[font = text; inactive; startGroup; Cclosed; ]
> And new Image is obtained by using Deconvolution Algorithm.
> This is case of no full reconstraction. It is determined by contact Tip with Surface do not all points.
> Otherwise will be seen a full reconstruction and new Image will be identical to initial Image.
>
> :[font = input; endGroup; nowordwrap; backColorRed = 65280; backColorGreen = 65280; backColorBlue = 65280; fontColorR
ed
> = 0; fontColorGreen = 0; fontColorBlue = 0; bold; fontName = "Courier New"; fontSize = 12; ]
> Table[Samp[[i,j]]=0,{i,Ny},{j,Nx}];
> Table[Samp[[i,j]]=New[[i-Nt,j-Nt]],{i,Nt+1,Ny-Nt},{j,Nt+1,Nx-Nt}];
> Do[New[[i,j]]=Min[N[Tip-Samp[[Range[i,2 Nt+i],Range[j,2Nt+j]]]]],
> {i,Ny-2 Nt},{j,Nx-2 Nt}];
> l2:=Table[GrayLevel[1],{i,Ny-2 Nt-1},{j,Nx-2 Nt-1}];
> q1=ListPlot3D[30-Max[New]+New,l2,PlotRange->{All,All,{-5,40}}];
> :[font = help; inactive; nohscroll; ]
> For question and more information You should be adressed :
> :[font = name; inactive; nohscroll; ]
> Kazan Physical Technical Insitute
> Nikolai Berdunov
> e-mail: nik at ksc.iasnet.ru
> or Nikolai.Berdunov at ksu.ru
> ^*)$
>
>
==== [MESSAGE SEPARATOR] ====