scikit-image October 2016

scikit-image@python.org

20 participants
31 discussions

Re: [scikit-image] BPDL in skimage?
by Stefan van der Walt 13 Nov '16

13 Nov '16

Hi Jiri On Mon, Oct 10, 2016, at 04:42, Jiří Borovec wrote: > recently we made a new method Binary Pattern Dictionary learning for > estimation an atlas from set of binary images where we detect and > decompose frequent patterns in the image set, see paper in attachment. > I was wondering if you are interested to have this method as > contribution to the scikit-image library? We typically wait a while for papers to gain popularity before we implement them in scikit-image. That said, I think it's an excellent idea to make the code available with proper licensing and documentation until such time that we can include it. Best regards Stéfan

3 3

Re: [scikit-image] GUI Python Image Processing Scikit-image
by Stefan van der Walt 31 Oct '16

31 Oct '16

This reminds me that we have a whole GUI viewer toolkit that's not seen a lot of love and attention lately. I recently ported it to Qt5, but still have to submit cleaned up patches. That got me thinking that it would be nice to have this in a browser, in a Jupyter notebook. So I sent an email to the developers, and it looks like they've now more-or- less completed their widgets API, and it should be mostly backward compatible from version 6 onward. Shall we re-open the discussion on writing interactive Javascript widgets? I'm thinking of a Fiji or Gimp-like application that showcases various pieces of functionality. Stéfan On Tue, Oct 25, 2016, at 15:43, Juan Nunez-Iglesias wrote: > Hi Xavi, > > Sorry for the long delay in responding. > > I'm excited about your GUI! If you have it on GitHub somewhere, > please point us to it, as it would be a great template for others to > build from! > > Regarding your black image problem, my guess is that you are having > some issues with the data types in skimage. Your input image is a > uint8 image (?), with range [0, 255], but filters.gaussian runs it > through our conversion to a float image, which results in the range > [0.0, 1.0]. Some displays will still display this as though the full > range is in [0, 255], so the image will appear black. You probably > need to back-convert to int using skimage.img_as_ubyte. > > See: > http://scikit-image.org/docs/dev/user_guide/data_types.html > > I hope that helps. Again, please point us to your repo if you are > happy to share! > > Juan. > > > On 8 October 2016 at 8:12:13 am, xavi barrull > (xavi_barrull(a)hotmail.com) wrote: >> >> >> >> Hi, I have a question about scikit image. >> I make a GUI in python to processing image, but I will change the >> gaussian Filter in opencv to scikit image, I use this function: >> >> def gaussFilter(self): targetTemp = >> sk.io.imread(self.original_path)# llegir imatge targetTemp a la ruta >> original targetTemp = sk.filters.gaussian(targetTemp, >> sigma=0, mode='reflect', cval=0, multichannel=False) >> #targetTemp = cv2.GaussianBlur(targetTemp,(sigmaX, sigmaY), 0) >> cv2.imwrite(self.target_path,targetTemp) >> self.updateTargetImage() >> >> But, when I press the button of gaussian filter the image to update >> becomes black, what is the problem? I need a solution for this! >> >> My code show bellow: >> >> >> import Tkinter as tk import numpy as np import tkFileDialog import >> tkMessageBox import shutil import os from PIL import Image, ImageTk, >> ImageFilter import cv2 import colormaps import skimage as sk import >> matplotlib.pyplot as plt from matplotlib.colors import ListedColormap >> from matplotlib.figure import Figure from skimage import filters from >> skimage import io from skimage.morphology import disk from >> skimage.filters.rank import median from skimage.filters import >> gaussian from skimage.morphology import watershed from >> matplotlib.backends.backend_tkagg import FigureCanvasTkAgg >> >> class App(tk.Frame): size = 500, 700 >> # get past scale value >> past=1 >> # inicialitzar app >> def __init__(self, master): tk.Frame.__init__(self, master) >> self.grid(row=0) self.columnconfigure(0, weight=1) >> self.rowconfigure(0, weight=1) master.geometry("1024x768") >> self.generateMenu(master) >> # Generacio Menu >> def generateMenu(self, master): >> # crear barra menu, dins de la llibreria TKinter (tk.) >> self.menuBar=tk.Menu(self) >> # creacio arxiu dins de la barra de menu >> self.fileMenu=tk.Menu(self.menuBar, tearoff=0) >> # crear commandes obrir, guardar i tancar, l'estat de >> # guardar esta desactivat mentre no s'obri una imatge >> self.fileMenu.add_command(label="Open", command=self.open) >> self.fileMenu.add_command(label="Save", command=self.save, >> state="disabled") self.fileMenu.add_command(label="Close", >> command=self.exit) >> # Afegim tot aixo la barra de menu en cascada >> self.menuBar.add_cascade(label="File", menu=self.fileMenu) >> # Crear menu de process >> self.processMenu=tk.Menu(self.menuBar, tearoff=0) >> # crear submenu de processs/filter >> self.processFilterSubMenu = tk.Menu(self.processMenu, >> tearoff=0) >> self.processFilterSubMenu.add_command(label="Gaussian >> Filter", command=self.gaussFilter, state="disabled") >> self.processFilterSubMenu.add_command(label="Median Filter", >> command=self.medianFilter, state="disabled") >> self.processMenu.add_cascade(label="Process/Filter", >> menu=self.processFilterSubMenu, underline=0) >> # crear submenu segmentacio >> self.segmentationSubMenu = tk.Menu(self.processMenu, >> tearoff=0) >> self.segmentationSubMenu.add_command(label="Watershed", >> command=self.watershed, state="disabled") >> self.segmentationSubMenu.add_command(label="Random walker", >> command=self.randomwalker, state="disabled") >> self.processMenu.add_cascade(label="Segmentation", >> menu=self.segmentationSubMenu, underline=0) >> # crear submenu edgeDetection >> self.edgeDetectionSubMenu = tk.Menu(self.processMenu, >> tearoff=0) >> self.edgeDetectionSubMenu.add_command(label="Sobel", >> command=self.sobel, state="disabled") >> self.edgeDetectionSubMenu.add_command(label="Laplacian", >> command=self.laplacian, state="disabled") >> self.edgeDetectionSubMenu.add_command(label="Canny", >> command=self.canny, state="disabled") >> self.processMenu.add_cascade(label="Edge Detection", >> menu=self.edgeDetectionSubMenu, underline=0) >> # crear submenu change colormap >> self.colorMapSubMenu = tk.Menu(self.processMenu, tearoff=0) >> self.colorMapSubMenu.add_command(label="None", >> command=self.none, state="disabled") >> self.colorMapSubMenu.add_command(label="Plasma", >> command=self.plasma, state="disabled") >> self.colorMapSubMenu.add_command(label="Inferno", >> command=self.inferno, state="disabled") >> self.colorMapSubMenu.add_command(label="Viridis", >> command=self.viridis, state="disabled") >> self.processMenu.add_cascade(label="Change Colormap", >> menu=self.colorMapSubMenu, underline=0) >> >> self.menuBar.add_cascade(label="Process", >> menu=self.processMenu) master.config(menu=self.menuBar) >> # No colorMap >> def none(self): targetTemp = cv2.imread(self.original_path) >> cv2.imwrite(self.target_path,targetTemp)# guarda imatge >> targetTemp i a la ruta de desti self.updateTargetImage() >> # Plasma colorMap >> def plasma(self): plt.clf()#esborra imatge targetTemp >> targetTemp = cv2.imread(self.original_path) # llegir ruta >> original (carpeta on es troba arxiu py) targetTemp = >> np.fliplr(targetTemp[:,:,0]) # gir esquerra a dreta imatge >> original, segons 3r valor matriu, varia tonalitat >> >> plasma = ListedColormap(colormaps._plasma_data, >> name='plasma') plt.axis('off')#desactiva eixos a la imatge >> plt.register_cmap(name='plasma', cmap=plasma) # registrar >> colormap nou i aplicar-lo imgplot = >> plt.imshow(targetTemp)#mostrar imatge allotjada a targetTemp >> imgplot.set_cmap(plasma)#aplicar colormap >> imgplot.figure.savefig(self.target_path)#guardar imatge com >> a figura a la ruta de desti >> >> self.updateTargetImage() #actualitzar la imatge (amb >> aplicacio del colormap) def viridis(self): plt.clf() >> #esborra imatge targetTemp targetTemp = >> cv2.imread(self.original_path) targetTemp = >> np.fliplr(targetTemp[:,:,0]) >> >> viridis = ListedColormap(colormaps._viridis_data, >> name='viridis') plt.axis('off') >> plt.register_cmap(name='viridis', cmap=viridis) imgplot = >> plt.imshow(targetTemp) imgplot.set_cmap(viridis) >> imgplot.figure.savefig(self.target_path) >> # update target screen >> self.updateTargetImage() def inferno(self): plt.clf() >> targetTemp = cv2.imread(self.original_path) targetTemp = >> np.fliplr(targetTemp[:,:,0]) >> >> inferno = ListedColormap(colormaps._inferno_data, >> name='inferno') plt.axis('off') >> plt.register_cmap(name='inferno', cmap=inferno) imgplot = >> plt.imshow(targetTemp) imgplot.set_cmap(inferno) >> imgplot.figure.savefig(self.target_path) >> self.updateTargetImage() >> >> # funcio per obrir arxius >> def open(self): >> # obrir arxiu des del quadre de dialeg >> self.original_path = >> tkFileDialog.askopenfilename(filetypes=[("Image >> Files","*.jpg;*.jpeg;*.png;*.gif")]) if >> self.original_path!="": # si es selecciona un arxiu, >> s'activen totes les comandes del menu >> self.fileMenu.entryconfig("Save", state="normal") >> self.processFilterSubMenu.entryconfig("Gaussian Filter", >> state="normal") >> self.processFilterSubMenu.entryconfig("Median Filter", >> state="normal") >> self.segmentationSubMenu.entryconfig("Watershed", >> state="normal") >> self.segmentationSubMenu.entryconfig("Random walker", >> state="normal") >> self.edgeDetectionSubMenu.entryconfig("Sobel", >> state="normal") >> self.edgeDetectionSubMenu.entryconfig("Laplacian", >> state="normal") >> self.edgeDetectionSubMenu.entryconfig("Canny", >> state="normal") self.colorMapSubMenu.entryconfig("None", >> state="normal") >> self.colorMapSubMenu.entryconfig("Plasma", state="normal") >> self.colorMapSubMenu.entryconfig("Inferno", state="normal") >> self.colorMapSubMenu.entryconfig("Viridis", state="normal") >> >> saveButton = tk.Button(self, text="Save", >> command=self.save) saveButton.grid(row=7, column=10) >> self.actual_name = self.original_path[self.original_pat- >> h.rfind("/"):self.original_path.rfind(".")] #creacio >> arxiu temporal de la imatge que hem obert >> self.target_path = self.original_path[0:self.original_p- >> ath.rfind(".")]+"imatge."+self.original_path[self.origi- >> nal_path.rfind(".")+1:] shutil.copy2(self.original_path, >> self.target_path) #copia l'arxiu de la ruta original >> #os.popen('attrib +h ' + self.target_path) >> self.originalImg = Image.open(self.original_path) #obrir >> imatge original self.originalImg.thumbnail(App.size, >> Image.ANTIALIAS) #thumbnail-> aplica mida imatge >> escalada a la finestra App.size originalImgCanvas = >> tk.Canvas(self, width= 500, height=380) # aplica la mida >> especificada a la imatge original >> originalImgCanvas.grid(row=3, column=2, columnspan=6) # >> la posiciona dins del grid a la fila 3 i la columna 2, >> abasta 6 columnes tkOriginalImg = >> ImageTk.PhotoImage(self.originalImg) # mostra la imatge >> original self.originalImg.image = tkOriginalImg # >> associa la imatge original a la imatge dins la llibreria >> tkinter originalImgCanvas.create_image(100,100,image=tk- >> OriginalImg, anchor=tk.NW, tags="IMG") # crea la imatge >> anterior amb amplada nord-oest dins del widget de Canvas >> que disposa d'estructures grafiques >> >> self.targetImg = Image.open(self.target_path)# obrir >> imatge a procesar self.targetImg.thumbnail(App.size, >> Image.ANTIALIAS) #thumbnail-> aplica mida imatge >> escalada a la finestra App.size self.targetImgCanvas = >> tk.Canvas(self, width= 500, height=380) >> self.targetImgCanvas.grid(row=3, column=8, columnspan=6) >> self.tkTargetImg = ImageTk.PhotoImage(self.targetImg) >> self.targetImg.image = self.tkTargetImg self.targetImgC- >> anvas.create_image(100,100,image=self.tkTargetImg, >> anchor=tk.NW) >> # source histogram >> #originalCvImg = cv2.imread(self.original_path) >> #plt.figure(figsize=(2,2)) #color = ('b','g','r') #for >> channel,col in enumerate(color): >> # histr = cv2.calcHist([originalCvImg],[channel],Non- >> # e,[256],[0,256]) plt.plot(histr,color = col) >> # plt.xlim([0,256]) if( col == 'b'): >> # originalB=plt.imshow(histr).figure elif( col == >> # 'g'): originalG=plt.imshow(histr).figure else: >> # originalR=plt.imshow(histr).figure >> #self.NewCanvas = tk.Canvas(self, width= 30, height=50) >> #self.canvas = FigureCanvasTkAgg(originalR, >> master=self.NewCanvas) #self.NewCanvas.grid(row=4, >> column=0) #self.canvas.show() >> #self.canvas.get_tk_widget().grid(row=0, column=0) >> >> # target histogram >> #targetCvImg = cv2.imread(self.target_path) >> #plt.figure(figsize=(2,2)) #color = ('b','g','r') #for >> channel,col in enumerate(color): >> # histr = cv2.calcHist([targetCvImg],[channel],None,[- >> # 256],[0,256]) plt.plot(histr,color = col) >> # plt.xlim([0,256]) if( col == 'b'): >> # targetB=plt.imshow(histr).figure elif( col == 'g'): >> # targetG=plt.imshow(histr).figure else: >> # targetR=plt.imshow(histr).figure >> #self.NewTargetCanvas = tk.Canvas(self, width= 30, >> height=50) #self.canvasTarget = >> FigureCanvasTkAgg(targetR, master=self.NewTargetCanvas) >> #self.NewTargetCanvas.grid(row=4, column=1) >> #self.canvasTarget.show() >> #self.canvasTarget.get_tk_widget().grid(row=0, column=0) >> # to have only odd value in scale >> def fix(self, xy, n): #no entenc que fa aquesta funcio! >> global past n = int(n) if not n % 2: if xy=="sx": >> scale = self.gaussianSigmaXScale elif xy=="sy": >> scale = self.gaussianSigmaYScale elif xy=="median": >> scale = self.medianSizeScale else: scale = >> self.kernelSizeScale scale.set(n+1 if n > self.past else >> n-1) self.past = scale.get() >> # funcio per guardar imatges procesades >> def save(self): if tkMessageBox.askokcancel("Save", "Do you >> want to save?"): os.remove(self.original_path)# esborra >> arxiu original de la ruta de desti os.popen('attrib -h -s >> ' + self.target_path)# obre arxiu target de >> # la ruta de desti, correspon imatge a processar >> os.rename(self.target_path, self.original_path) >> #renombra >> # la imatge original i la imatge a processar >> shutil.copy2(self.original_path, self.target_path) >> #copia l'arxiu >> # de la ruta original i tambe la imatge a processar >> # (target) >> os.popen('attrib +h ' + self.target_path) >> # sortir del GUI >> def exit(self): if tkMessageBox.askokcancel("Quit", "Do you >> want to quit?"): try: >> # esborrar arxiu temporal >> os.remove(self.target_path) finally: >> self.master.destroy() >> >> def updateTargetImage(self): # funcio per actualitzar les >> diferents accio a la imatge a procesar self.targetImg = >> Image.open(self.target_path) # obre la imatge a procesar, en la >> ruta on es troba self.targetImg.thumbnail(App.size, >> Image.ANTIALIAS) #thumbnail-> aplica mida imatge escalada a la >> finestra App.size self.tkTargetImg = >> ImageTk.PhotoImage(self.targetImg)# mostra la imatge a procesar >> self.targetImg.image = self.tkTargetImg # associa la imatge a >> procesar a la imatge dins la llibreria Tkinter (que allotja tot >> el referent al GUI) self.targetImgCanvas.create_image(5,5,im- >> age=self.tkTargetImg, anchor=tk.NW, tags="IMG") # crea la >> #imatge a procesar quan se li aplica algun canvi (colormap, >> filtres, etc) targetCvImg = cv2.imread(self.target_path) # >> llegeix la imatge de la ruta de desti >> plt.figure(figsize=(4,4)) # crea una figura per allotjar >> histograma color = ('b','g','r') for i,col in >> enumerate(color): histr = >> cv2.calcHist([targetCvImg],[i],None,[256],[0,256]) >> plt.plot(histr,color = col) plt.xlim([0,256]) >> targetR=plt.imshow(histr).figure self.NewTargetCanvas = >> tk.Canvas(self, width= 30, height=50) #crea un nou >> # Canvas de la mida indicada >> self.canvasTarget = FigureCanvasTkAgg(targetR, >> master=self.NewTargetCanvas) #crea una figura dins del >> Canvas on s'allotja el histograma >> self.NewTargetCanvas.grid(row=4, column=1)# defineix posicio >> del grid dins el >> # Canvas >> self.canvasTarget.show() # mostra el canvas >> self.canvasTarget.get_tk_widget().grid(row=0, column=0) >> >> def gaussFilter(self): targetTemp = >> sk.io.imread(self.original_path)# llegir imatge targetTemp a la >> ruta original targetTemp = sk.filters.gaussian(targetTemp, >> sigma=0, mode='reflect', cval=0, multichannel=False) >> #targetTemp = cv2.GaussianBlur(targetTemp,(sigmaX, sigmaY), >> 0)#aplicar filtre cv2.imwrite(self.target_path,targetTemp) >> # actualitzar imatge Target >> self.updateTargetImage() >> >> My code is so long (There are another functions or filters (Sobel, >> Canny, etc...) I paste the finally: >> >> def on_closing(app, root): if tkMessageBox.askokcancel("Quit", >> "Do you want to quit?"): try: >> os.remove(app.target_path) except: print("error") >> finally: root.destroy() >> # inicialitzar llibreria Tkinter per creacio widget >> root=tk.Tk() app = App(root) >> # renderitzar GUI, to accept events >> root.protocol("WM_DELETE_WINDOW", lambda: on_closing(app, root)) >> app.mainloop() >> >> >> I attached image about my problema >> >> I need help, it's so urgent! >> >> Thanks in advance. >> >> Xavi. >> _______________________________________________ >> scikit-image mailing list scikit-image(a)python.org >> https://mail.python.org/mailman/listinfo/scikit-image >> > _________________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image

1 0

Re: [scikit-image] find grid points inside verts
by Slavin, Jonathan 31 Oct '16

31 Oct '16

Li, So are you saying to use a slice, get a contour, e.g. with get_contour, and then use grid_points_in_poly()? I suppose that could work for part of the volume, but I would question its effectiveness for my case (quasi-spherical surface in 3D cartesian coordinates) in regions where the slice and surface are close to parallel. If I could get surfaces that are, say phi=const. surfaces, that include the origin then that could work, but then that require an interpolation onto each grid, and to cover all the voxels, I'd need many surfaces. A different thought I had was to generate a ray using methods in yt, for each voxel and find the outermost position along the ray where the surface value is obtained. That could also be very slow, but for my case, I could limit it by only doing that calculation for voxels where the value (in my case pressure) is below the surface value, since all the voxels with pressure above that value, are inside the outer surface. It would be nice to have a more general solution if such a thing exists (which seems likely). Regards, Jon On Sun, Oct 23, 2016 at 9:15 PM, Lee Chen <lchenudc(a)gmail.com> wrote: > This is an interesting question. Can you try to use grid_points_in_poly() > for each slice of 3D. It may be slower. I think there is a mathematical > extension on this. But I do not know the actual formula. > > Li Chen > www.udc.edu/prof/chen > > > Does anyone know SciKit work with Spark (Python version)? > > > On Tue, Oct 18, 2016 at 1:57 PM, Slavin, Jonathan <jslavin(a)cfa.harvard.edu > > wrote: > >> Hi all, >> >> I'm wondering if there is some routine that will do in 3D what >> grid_points_in_poly() does in 2D. That is, I'd like to use the results of >> a call to marching_cubes(), i.e. an array of verts, and determine which >> grid points in the input volume that was provided as input to >> marching_cubes are inside of the surface defined by those verts. Any help >> would be appreciated. >> >> Regards, >> Jon >> >> -- >> ________________________________________________________ >> Jonathan D. Slavin Harvard-Smithsonian CfA >> jslavin(a)cfa.harvard.edu 60 Garden Street, MS 83 >> phone: (617) 496-7981 Cambridge, MA 02138-1516 >> cell: (781) 363-0035 USA >> ________________________________________________________ >> >> >> _______________________________________________ >> scikit-image mailing list >> scikit-image(a)python.org >> https://mail.python.org/mailman/listinfo/scikit-image >> >> > -- ________________________________________________________ Jonathan D. Slavin Harvard-Smithsonian CfA jslavin(a)cfa.harvard.edu 60 Garden Street, MS 83 phone: (617) 496-7981 Cambridge, MA 02138-1516 cell: (781) 363-0035 USA ________________________________________________________

1 0

Re: [scikit-image] GUI Python Image Processing Scikit-image
by Juan Nunez-Iglesias 25 Oct '16

25 Oct '16

Hi Xavi, Sorry for the long delay in responding. I'm excited about your GUI! If you have it on GitHub somewhere, please point us to it, as it would be a great template for others to build from! Regarding your black image problem, my guess is that you are having some issues with the data types in skimage. Your input image is a uint8 image (?), with range [0, 255], but filters.gaussian runs it through our conversion to a float image, which results in the range [0.0, 1.0]. Some displays will still display this as though the full range is in [0, 255], so the image will appear black. You probably need to back-convert to int using skimage.img_as_ubyte. See: http://scikit-image.org/docs/dev/user_guide/data_types.html I hope that helps. Again, please point us to your repo if you are happy to share! Juan. On 8 October 2016 at 8:12:13 am, xavi barrull (xavi_barrull(a)hotmail.com) wrote: Hi, I have a question about scikit image. I make a GUI in python to processing image, but I will change the gaussian Filter in opencv to scikit image, I use this function: def gaussFilter(self): targetTemp = sk.io.imread(self.original_path)# llegir imatge targetTemp a la ruta original targetTemp = sk.filters.gaussian(targetTemp, sigma=0, mode='reflect', cval=0, multichannel=False) #targetTemp = cv2.GaussianBlur(targetTemp,(sigmaX, sigmaY), 0) cv2.imwrite(self.target_path,targetTemp) self.updateTargetImage() But, when I press the button of gaussian filter the image to update becomes black, what is the problem? I need a solution for this! My code show bellow: import Tkinter as tk import numpy as np import tkFileDialog import tkMessageBox import shutil import os from PIL import Image, ImageTk, ImageFilter import cv2 import colormaps import skimage as sk import matplotlib.pyplot as plt from matplotlib.colors import ListedColormap from matplotlib.figure import Figure from skimage import filters from skimage import io from skimage.morphology import disk from skimage.filters.rank import median from skimage.filters import gaussian from skimage.morphology import watershed from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg class App(tk.Frame): size = 500, 700 # get past scale value past=1 # inicialitzar app def __init__(self, master): tk.Frame.__init__(self, master) self.grid(row=0) self.columnconfigure(0, weight=1) self.rowconfigure(0, weight=1) master.geometry("1024x768") self.generateMenu(master) # Generacio Menu def generateMenu(self, master): # crear barra menu, dins de la llibreria TKinter (tk.) self.menuBar=tk.Menu(self) # creacio arxiu dins de la barra de menu self.fileMenu=tk.Menu(self.menuBar, tearoff=0) # crear commandes obrir, guardar i tancar, l'estat de guardar esta desactivat mentre no s'obri una imatge self.fileMenu.add_command(label="Open", command=self.open) self.fileMenu.add_command(label="Save", command=self.save, state="disabled") self.fileMenu.add_command(label="Close", command=self.exit) # Afegim tot aixo la barra de menu en cascada self.menuBar.add_cascade(label="File", menu=self.fileMenu) # Crear menu de process self.processMenu=tk.Menu(self.menuBar, tearoff=0) # crear submenu de processs/filter self.processFilterSubMenu = tk.Menu(self.processMenu, tearoff=0) self.processFilterSubMenu.add_command(label="Gaussian Filter", command=self.gaussFilter, state="disabled") self.processFilterSubMenu.add_command(label="Median Filter", command=self.medianFilter, state="disabled") self.processMenu.add_cascade(label="Process/Filter", menu=self.processFilterSubMenu, underline=0) # crear submenu segmentacio self.segmentationSubMenu = tk.Menu(self.processMenu, tearoff=0) self.segmentationSubMenu.add_command(label="Watershed", command=self.watershed, state="disabled") self.segmentationSubMenu.add_command(label="Random walker", command=self.randomwalker, state="disabled") self.processMenu.add_cascade(label="Segmentation", menu=self.segmentationSubMenu, underline=0) # crear submenu edgeDetection self.edgeDetectionSubMenu = tk.Menu(self.processMenu, tearoff=0) self.edgeDetectionSubMenu.add_command(label="Sobel", command=self.sobel, state="disabled") self.edgeDetectionSubMenu.add_command(label="Laplacian", command=self.laplacian, state="disabled") self.edgeDetectionSubMenu.add_command(label="Canny", command=self.canny, state="disabled") self.processMenu.add_cascade(label="Edge Detection", menu=self.edgeDetectionSubMenu, underline=0) # crear submenu change colormap self.colorMapSubMenu = tk.Menu(self.processMenu, tearoff=0) self.colorMapSubMenu.add_command(label="None", command=self.none, state="disabled") self.colorMapSubMenu.add_command(label="Plasma", command=self.plasma, state="disabled") self.colorMapSubMenu.add_command(label="Inferno", command=self.inferno, state="disabled") self.colorMapSubMenu.add_command(label="Viridis", command=self.viridis, state="disabled") self.processMenu.add_cascade(label="Change Colormap", menu=self.colorMapSubMenu, underline=0) self.menuBar.add_cascade(label="Process", menu=self.processMenu) master.config(menu=self.menuBar) # No colorMap def none(self): targetTemp = cv2.imread(self.original_path) cv2.imwrite(self.target_path,targetTemp)# guarda imatge targetTemp i a la ruta de desti self.updateTargetImage() # Plasma colorMap def plasma(self): plt.clf()#esborra imatge targetTemp targetTemp = cv2.imread(self.original_path) # llegir ruta original (carpeta on es troba arxiu py) targetTemp = np.fliplr(targetTemp[:,:,0]) # gir esquerra a dreta imatge original, segons 3r valor matriu, varia tonalitat plasma = ListedColormap(colormaps._plasma_data, name='plasma') plt.axis('off')#desactiva eixos a la imatge plt.register_cmap(name='plasma', cmap=plasma) # registrar colormap nou i aplicar-lo imgplot = plt.imshow(targetTemp)#mostrar imatge allotjada a targetTemp imgplot.set_cmap(plasma)#aplicar colormap imgplot.figure.savefig(self.target_path)#guardar imatge com a figura a la ruta de desti self.updateTargetImage() #actualitzar la imatge (amb aplicacio del colormap) def viridis(self): plt.clf() #esborra imatge targetTemp targetTemp = cv2.imread(self.original_path) targetTemp = np.fliplr(targetTemp[:,:,0]) viridis = ListedColormap(colormaps._viridis_data, name='viridis') plt.axis('off') plt.register_cmap(name='viridis', cmap=viridis) imgplot = plt.imshow(targetTemp) imgplot.set_cmap(viridis) imgplot.figure.savefig(self.target_path) # update target screen self.updateTargetImage() def inferno(self): plt.clf() targetTemp = cv2.imread(self.original_path) targetTemp = np.fliplr(targetTemp[:,:,0]) inferno = ListedColormap(colormaps._inferno_data, name='inferno') plt.axis('off') plt.register_cmap(name='inferno', cmap=inferno) imgplot = plt.imshow(targetTemp) imgplot.set_cmap(inferno) imgplot.figure.savefig(self.target_path) self.updateTargetImage() # funcio per obrir arxius def open(self): # obrir arxiu des del quadre de dialeg self.original_path = tkFileDialog.askopenfilename(filetypes=[("Image Files","*.jpg;*.jpeg;*.png;*.gif")]) if self.original_path!="": # si es selecciona un arxiu, s'activen totes les comandes del menu self.fileMenu.entryconfig("Save", state="normal") self.processFilterSubMenu.entryconfig("Gaussian Filter", state="normal") self.processFilterSubMenu.entryconfig("Median Filter", state="normal") self.segmentationSubMenu.entryconfig("Watershed", state="normal") self.segmentationSubMenu.entryconfig("Random walker", state="normal") self.edgeDetectionSubMenu.entryconfig("Sobel", state="normal") self.edgeDetectionSubMenu.entryconfig("Laplacian", state="normal") self.edgeDetectionSubMenu.entryconfig("Canny", state="normal") self.colorMapSubMenu.entryconfig("None", state="normal") self.colorMapSubMenu.entryconfig("Plasma", state="normal") self.colorMapSubMenu.entryconfig("Inferno", state="normal") self.colorMapSubMenu.entryconfig("Viridis", state="normal") saveButton = tk.Button(self, text="Save", command=self.save) saveButton.grid(row=7, column=10) self.actual_name = self.original_path[self.original_path.rfind("/"):self.original_path.rfind(".")] #creacio arxiu temporal de la imatge que hem obert self.target_path = self.original_path[0:self.original_path.rfind(".")]+"imatge."+self.original_path[self.original_path.rfind(".")+1:] shutil.copy2(self.original_path, self.target_path) #copia l'arxiu de la ruta original #os.popen('attrib +h ' + self.target_path) self.originalImg = Image.open(self.original_path) #obrir imatge original self.originalImg.thumbnail(App.size, Image.ANTIALIAS) #thumbnail-> aplica mida imatge escalada a la finestra App.size originalImgCanvas = tk.Canvas(self, width= 500, height=380) # aplica la mida especificada a la imatge original originalImgCanvas.grid(row=3, column=2, columnspan=6) # la posiciona dins del grid a la fila 3 i la columna 2, abasta 6 columnes tkOriginalImg = ImageTk.PhotoImage(self.originalImg) # mostra la imatge original self.originalImg.image = tkOriginalImg # associa la imatge original a la imatge dins la llibreria tkinter originalImgCanvas.create_image(100,100,image=tkOriginalImg, anchor=tk.NW, tags="IMG") # crea la imatge anterior amb amplada nord-oest dins del widget de Canvas que disposa d'estructures grafiques self.targetImg = Image.open(self.target_path)# obrir imatge a procesar self.targetImg.thumbnail(App.size, Image.ANTIALIAS) #thumbnail-> aplica mida imatge escalada a la finestra App.size self.targetImgCanvas = tk.Canvas(self, width= 500, height=380) self.targetImgCanvas.grid(row=3, column=8, columnspan=6) self.tkTargetImg = ImageTk.PhotoImage(self.targetImg) self.targetImg.image = self.tkTargetImg self.targetImgCanvas.create_image(100,100,image=self.tkTargetImg, anchor=tk.NW) # source histogram #originalCvImg = cv2.imread(self.original_path) #plt.figure(figsize=(2,2)) #color = ('b','g','r') #for channel,col in enumerate(color): # histr = cv2.calcHist([originalCvImg],[channel],None,[256],[0,256]) # plt.plot(histr,color = col) # plt.xlim([0,256]) # if( col == 'b'): # originalB=plt.imshow(histr).figure # elif( col == 'g'): # originalG=plt.imshow(histr).figure # else: # originalR=plt.imshow(histr).figure #self.NewCanvas = tk.Canvas(self, width= 30, height=50) #self.canvas = FigureCanvasTkAgg(originalR, master=self.NewCanvas) #self.NewCanvas.grid(row=4, column=0) #self.canvas.show() #self.canvas.get_tk_widget().grid(row=0, column=0) # target histogram #targetCvImg = cv2.imread(self.target_path) #plt.figure(figsize=(2,2)) #color = ('b','g','r') #for channel,col in enumerate(color): # histr = cv2.calcHist([targetCvImg],[channel],None,[256],[0,256]) # plt.plot(histr,color = col) # plt.xlim([0,256]) # if( col == 'b'): # targetB=plt.imshow(histr).figure # elif( col == 'g'): # targetG=plt.imshow(histr).figure # else: # targetR=plt.imshow(histr).figure #self.NewTargetCanvas = tk.Canvas(self, width= 30, height=50) #self.canvasTarget = FigureCanvasTkAgg(targetR, master=self.NewTargetCanvas) #self.NewTargetCanvas.grid(row=4, column=1) #self.canvasTarget.show() #self.canvasTarget.get_tk_widget().grid(row=0, column=0) # to have only odd value in scale def fix(self, xy, n): #no entenc que fa aquesta funcio! global past n = int(n) if not n % 2: if xy=="sx": scale = self.gaussianSigmaXScale elif xy=="sy": scale = self.gaussianSigmaYScale elif xy=="median": scale = self.medianSizeScale else: scale = self.kernelSizeScale scale.set(n+1 if n > self.past else n-1) self.past = scale.get() # funcio per guardar imatges procesades def save(self): if tkMessageBox.askokcancel("Save", "Do you want to save?"): os.remove(self.original_path)# esborra arxiu original de la ruta de desti os.popen('attrib -h -s ' + self.target_path)# obre arxiu target de # la ruta de desti, correspon imatge a processar os.rename(self.target_path, self.original_path) #renombra # la imatge original i la imatge a processar shutil.copy2(self.original_path, self.target_path) #copia l'arxiu # de la ruta original i tambe la imatge a processar (target) os.popen('attrib +h ' + self.target_path) # sortir del GUI def exit(self): if tkMessageBox.askokcancel("Quit", "Do you want to quit?"): try: # esborrar arxiu temporal os.remove(self.target_path) finally: self.master.destroy() def updateTargetImage(self): # funcio per actualitzar les diferents accio a la imatge a procesar self.targetImg = Image.open(self.target_path) # obre la imatge a procesar, en la ruta on es troba self.targetImg.thumbnail(App.size, Image.ANTIALIAS) #thumbnail-> aplica mida imatge escalada a la finestra App.size self.tkTargetImg = ImageTk.PhotoImage(self.targetImg)# mostra la imatge a procesar self.targetImg.image = self.tkTargetImg # associa la imatge a procesar a la imatge dins la llibreria Tkinter (que allotja tot el referent al GUI) self.targetImgCanvas.create_image(5,5,image=self.tkTargetImg, anchor=tk.NW, tags="IMG") # crea la #imatge a procesar quan se li aplica algun canvi (colormap, filtres, etc) targetCvImg = cv2.imread(self.target_path) # llegeix la imatge de la ruta de desti plt.figure(figsize=(4,4)) # crea una figura per allotjar histograma color = ('b','g','r') for i,col in enumerate(color): histr = cv2.calcHist([targetCvImg],[i],None,[256],[0,256]) plt.plot(histr,color = col) plt.xlim([0,256]) targetR=plt.imshow(histr).figure self.NewTargetCanvas = tk.Canvas(self, width= 30, height=50) #crea un nou # Canvas de la mida indicada self.canvasTarget = FigureCanvasTkAgg(targetR, master=self.NewTargetCanvas) #crea una figura dins del Canvas on s'allotja el histograma self.NewTargetCanvas.grid(row=4, column=1)# defineix posicio del grid dins el # Canvas self.canvasTarget.show() # mostra el canvas self.canvasTarget.get_tk_widget().grid(row=0, column=0) def gaussFilter(self): targetTemp = sk.io.imread(self.original_path)# llegir imatge targetTemp a la ruta original targetTemp = sk.filters.gaussian(targetTemp, sigma=0, mode='reflect', cval=0, multichannel=False) #targetTemp = cv2.GaussianBlur(targetTemp,(sigmaX, sigmaY), 0)#aplicar filtre cv2.imwrite(self.target_path,targetTemp) # actualitzar imatge Target self.updateTargetImage() My code is so long (There are another functions or filters (Sobel, Canny, etc...) I paste the finally: def on_closing(app, root): if tkMessageBox.askokcancel("Quit", "Do you want to quit?"): try: os.remove(app.target_path) except: print("error") finally: root.destroy() # inicialitzar llibreria Tkinter per creacio widget root=tk.Tk() app = App(root) # renderitzar GUI, to accept events root.protocol("WM_DELETE_WINDOW", lambda: on_closing(app, root)) app.mainloop() I attached image about my problema I need help, it's so urgent! Thanks in advance. Xavi. _______________________________________________ scikit-image mailing list scikit-image(a)python.org https://mail.python.org/mailman/listinfo/scikit-image

1 0

Re: [scikit-image] find grid points inside verts
by Lee Chen 24 Oct '16

24 Oct '16

This is an interesting question. Can you try to use grid_points_in_poly() for each slice of 3D. It may be slower. I think there is a mathematical extension on this. But I do not know the actual formula. Li Chen www.udc.edu/prof/chen Does anyone know SciKit work with Spark (Python version)? On Tue, Oct 18, 2016 at 1:57 PM, Slavin, Jonathan <jslavin(a)cfa.harvard.edu> wrote: > Hi all, > > I'm wondering if there is some routine that will do in 3D what > grid_points_in_poly() does in 2D. That is, I'd like to use the results of > a call to marching_cubes(), i.e. an array of verts, and determine which > grid points in the input volume that was provided as input to > marching_cubes are inside of the surface defined by those verts. Any help > would be appreciated. > > Regards, > Jon > > -- > ________________________________________________________ > Jonathan D. Slavin Harvard-Smithsonian CfA > jslavin(a)cfa.harvard.edu 60 Garden Street, MS 83 > phone: (617) 496-7981 Cambridge, MA 02138-1516 > cell: (781) 363-0035 USA > ________________________________________________________ > > > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image > >

1 0

find grid points inside verts
by Slavin, Jonathan 18 Oct '16

18 Oct '16

Hi all, I'm wondering if there is some routine that will do in 3D what grid_points_in_poly() does in 2D. That is, I'd like to use the results of a call to marching_cubes(), i.e. an array of verts, and determine which grid points in the input volume that was provided as input to marching_cubes are inside of the surface defined by those verts. Any help would be appreciated. Regards, Jon -- ________________________________________________________ Jonathan D. Slavin Harvard-Smithsonian CfA jslavin(a)cfa.harvard.edu 60 Garden Street, MS 83 phone: (617) 496-7981 Cambridge, MA 02138-1516 cell: (781) 363-0035 USA ________________________________________________________

1 0

Re: [scikit-image] Fast Radon Transform contributing
by Thomas Caswell 14 Oct '16

14 Oct '16

It would also be good to touch base with the tomopy developers ( https://tomopy.github.io/tomopy/resources/tomopy.algorithms.recon.html) Tom On Wed, Oct 12, 2016 at 6:16 PM Nadim Farhat <nadim.farhat(a)gmail.com> wrote: > Hi , > > That sounds great , Thanks > > Bests > > Nadim > > > On Wed, Oct 12, 2016 at 1:37 PM Stefan van der Walt <stefanv(a)berkeley.edu> > wrote: > > Hi Nadim > > On Wed, Oct 12, 2016, at 09:31, Nadim Farhat wrote: > > I am Nadim Farhat, Ph.D. Bioengineering Student at the University of > Pittsburgh. > I am writing this email to introduce myself as a future contributor also > to ask about the status of the Fast Radon Transform that I found from > https://github.com/scikit-image/scikit-image/wiki/Requested-features. I > would like to implement this feature in scikit-image. Please let me know > any feedback or issues. > > > It's good to meet you, and welcome! I would be happy to put you in touch > with others at LBNL doing this type of reconstruction; their perspectives > may be helpful. We can also look at adding SIRT (in addition to SART, > that's already there). > > Best regards > Stéfan > > > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image > > -- > Nadim Farhat > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image >

1 0

Re: [scikit-image] Fast Radon Transform contributing
by Nadim Farhat 12 Oct '16

12 Oct '16

Hi , That sounds great , Thanks Bests Nadim On Wed, Oct 12, 2016 at 1:37 PM Stefan van der Walt <stefanv(a)berkeley.edu> wrote: Hi Nadim On Wed, Oct 12, 2016, at 09:31, Nadim Farhat wrote: I am Nadim Farhat, Ph.D. Bioengineering Student at the University of Pittsburgh. I am writing this email to introduce myself as a future contributor also to ask about the status of the Fast Radon Transform that I found from https://github.com/scikit-image/scikit-image/wiki/Requested-features. I would like to implement this feature in scikit-image. Please let me know any feedback or issues. It's good to meet you, and welcome! I would be happy to put you in touch with others at LBNL doing this type of reconstruction; their perspectives may be helpful. We can also look at adding SIRT (in addition to SART, that's already there). Best regards Stéfan _______________________________________________ scikit-image mailing list scikit-image(a)python.org https://mail.python.org/mailman/listinfo/scikit-image -- Nadim Farhat

1 0

Re: [scikit-image] Fast Radon Transform contributing
by Ferhat KURTULMUŞ 12 Oct '16

12 Oct '16

I love scikit-image but want to unsubscribe from the list. Thanks in advance for any help. On Wed, Oct 12, 2016 at 8:37 PM, Stefan van der Walt <stefanv(a)berkeley.edu> wrote: > Hi Nadim > > On Wed, Oct 12, 2016, at 09:31, Nadim Farhat wrote: > > I am Nadim Farhat, Ph.D. Bioengineering Student at the University of > Pittsburgh. > I am writing this email to introduce myself as a future contributor also > to ask about the status of the Fast Radon Transform that I found from > https://github.com/scikit-image/scikit-image/wiki/Requested-features. I > would like to implement this feature in scikit-image. Please let me know > any feedback or issues. > > > It's good to meet you, and welcome! I would be happy to put you in touch > with others at LBNL doing this type of reconstruction; their perspectives > may be helpful. We can also look at adding SIRT (in addition to SART, > that's already there). > > Best regards > Stéfan > > > > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image > >

1 0

Re: [scikit-image] Fast Radon Transform contributing
by Stefan van der Walt 12 Oct '16

12 Oct '16

Hi Nadim On Wed, Oct 12, 2016, at 09:31, Nadim Farhat wrote: > I am Nadim Farhat, Ph.D. Bioengineering Student at the University of > Pittsburgh. > I am writing this email to introduce myself as a future contributor > also to ask about the status of the Fast Radon Transform that I found > from > https://github.com/scikit-image/scikit-image/wiki/Requested-features. > I would like to implement this feature in scikit-image. Please let me > know any feedback or issues. It's good to meet you, and welcome! I would be happy to put you in touch with others at LBNL doing this type of reconstruction; their perspectives may be helpful. We can also look at adding SIRT (in addition to SART, that's already there). Best regards Stéfan

1 0

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

scikit-image October 2016