Mailman 3 November 2015 - scikit-image

Counting objects using label in skimage.measure
by Arctic_python 19 Nov '15

19 Nov '15

Hello, I am trying to count the number of objects(that has "ones") in a binary array. Of course I do not expect to count holes. But when I run the below code, file1='4.csv' a=np.loadtxt(open(file1,'rb'),delimiter=',',dtype=int) #print (a.shape) img=measure.label(a) propsa = measure.regionprops(img) length = len(propsa) print ('length='+str(length)) for label in propsa: print (label.centroid) returns length=2 (214.23444957510378, 505.25546156532539) (238.77173913043478, 740.28260869565213) I get two objects. From reading the centroid coordinates(above), it seems it is counting the center of the white object and the cavity (can be seen in the image bellow). Why is this algorithm counting cavities and not only objects that are of "ones"? Is there an argument to enforce only objects and not cavities? Attached is the csv file if you want to try for yourself. Thanks

2 2

Implementation of imtool for skimage
by Darleison Rodrigues 19 Nov '15

19 Nov '15

Someone developing this? i will start a project to develop a lot of image functions from MATLAB and i want know who is working in this field.

2 1

Re: measuring the longest thread of a skeleton
by Juan Nunez-Iglesias 19 Nov '15

19 Nov '15

Eyal, not naive at all, it's a reasonable expectation! Skeletonization and related algorithms are a place where the library needs improving. If you come up with a good solution, we can help you get it into the library. See http://scikit-image.org/docs/stable/contribute.html . Thanks! Juan. On Thu, Nov 19, 2015 at 7:07 AM, Eyal Saiet <ejsaiet(a)alaska.edu> wrote: > Thanks I will look into the network approach. > I guess I was naive to assume there is a common algorithm, built in > scikit-image to measure the length of the skeleton. > > On Tue, Nov 17, 2015 at 9:43 PM, Pratap Vardhan <pratapgr8(a)gmail.com> > wrote: > >> My first thought was what Juan suggested and seemed logical to do that. >> >> As an alternative, you could also (this may be an overfill and could be >> slower than network approach) try. >> >> 1. From every endpoints of skeleton compute the distance transform (using >> flood-fill or neighbourhood methods). >> 2. Now the maximum distance for above all distances will give you >> the longest path in skeleton. >> >> This way you can have the trace path of the longest thread in skeleton in >> image form itself. >> >> >> On Wednesday, November 18, 2015 at 8:52:05 AM UTC+5:30, Arctic_python >> wrote: >>> >>> Hello, >>> Anyone has suggestions for an algorithm to measure the length of a >>> skeleton line/thread(e.g >>> http://scikit-image.org/docs/dev/auto_examples/plot_skeleton.html)? >>> The context- I skeletonize a shape to infer its length. >>> Thanks >>> >> -- >> You received this message because you are subscribed to a topic in the >> Google Groups "scikit-image" group. >> To unsubscribe from this topic, visit >> https://groups.google.com/d/topic/scikit-image/wM-zMGL9dVI/unsubscribe. >> To unsubscribe from this group and all its topics, send an email to >> scikit-image+unsubscribe(a)googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. >> > > > > -- > Eyal Saiet > > Project manager > Remote sensing and in-situ measurements > > Geophysical Institute > University of Alaska Fairbanks > Fairbanks, AK 99775 > (907) 750 6555 (cell) > > ejsaiet(a)alaska.edu > > -- > You received this message because you are subscribed to the Google Groups > "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/d/optout. >

1 0

measuring the longest thread of a skeleton
by Arctic_python 18 Nov '15

18 Nov '15

Hello, Anyone has suggestions for an algorithm to measure the length of a skeleton line/thread(e.g http://scikit-image.org/docs/dev/auto_examples/plot_skeleton.html)? The context- I skeletonize a shape to infer its length. Thanks

5 4

Improving HoG
by Martin Savc 17 Nov '15

17 Nov '15

I've been implementing my own HoG transform looking at different sources. While the implementation in scikit-image seems to lack certain features (multiple normalization schemes, general block overlap, Gaussian block window, trillinear interpolation/weighting of bin assignments,...) these don't seem to be that important, at least when applied to my current problem (eye blink analysis). Most of these would increase complexity, giving the implementation a complicated look, with little gain. I've also been looking into some practical improvements - integral histogram, separating the cell-block histogram feature to use it with other dense feature transforms such as LBP, a HoG visualization function that would render the visualization at higher resolutions that the original image. Would any of these be welcomed additions to scikit-image? Regards, Martin Savc PuppySaturation

3 3

Problems with coordinate ranges when doing conversion from LCH to LAB to RGB
by Matteo 16 Nov '15

16 Nov '15

A while ago I wrote a post on my blog with code on how to import in Python a colour palette in an ASCII file and convert it to a Matplotlib colormap: mycarta.wordpress.com/2014/04/25/convert-color-palettes-to-python-matplotli… Following that post I wrote a tutorial on a geoscience magazine on how to evaluate and compare colormaps using Python: wiki.seg.org/wiki/How_to_evaluate_and_compare_color_maps In the accompanying notebook I show how to convert a 256x3 RGB colormap to a 256x256x3 RGB image, then convert it to CIELAB using scikit-image's rgb2lab, then plot the 256x1 lightness (L) array to evaluate the colormaps perceptually. You can read the relevant extract of the notebook using this nbviewer link: http://nbviewer.ipython.org/urls/dl.dropbox.com/s/54zoax2qesb71wn/evaluate_… In that case scikit-image worked really well for me. Now I am trying to follow up with a new tutorial and I run into problems with the color space conversions. You can follow what I am trying to do in this other notebook extract: http://nbviewer.ipython.org/urls/dl.dropbox.com/s/noli66nzrlk0676/make_colo… The goal of this new tutorial is to show how to build colormaps from scratch using perceptual principles. I design a color palette in LCH (polar version of CIELAB) by keeping Chroma and Lightness fixed and interpolating Hue around the circle, then convert to LAB, then to RGB. As far as I know the code I wrote should work, but the result is a black colormap. I am thinking I got wrong one or more of the ranges for the LCH coordinates. I assumed L between (0,1), Ch between (0,1), and H between (0, 2*pi). Is that wrong, and if that's the case, what are the ranges? Many of them are not stated clearly in the documentation in here: http://scikit-image.org/docs/dev/api/skimage.color.html Is it possible to update the documentation to clearly state all ranges for all colour spaces. Thanks for your help. Matteo

3 5

Issue with morphological filters
by Matteo 16 Nov '15

16 Nov '15

*Issues with morphological filters when trying to remove white holes in black objects in a binary images. Using opening or filling holes on inverted (or complement) of the original binary.* Hi there I have a series of derivatives calculated on geophysical data. Many of these derivatives have nice continuous maxima, so I treat them as images on which I do some cleanup with morphological filter. Here's one example of operations that I do routinely, and successfully: # threshold theta map using Otsu method thresh_th = threshold_otsu(theta) binary_th = theta > thresh_th # clean up small objects label_objects_th, nb_labels_th = sp.ndimage.label(binary_th) sizes_th = np.bincount(label_objects_th.ravel()) mask_sizes_th = sizes_th > 175 mask_sizes_th[0] = 0 binary_cleaned_th = mask_sizes_th[label_objects_th] # further enhance with morphological closing (dilation followed by an erosion) to remove small dark spots and connect small bright cracks # followed by an extra erosion selem = disk(1) closed_th = closing(binary_cleaned_th, selem)/255 eroded_th = erosion(closed_th, selem)/255 # Finally, extract lienaments using skeletonization skeleton_th=skeletonize(binary_th) skeleton_cleaned_th=skeletonize(binary_cleaned_th) # plot to compare fig = plt.figure(figsize=(20, 7)) ax = fig.add_subplot(1, 2, 1) imshow(skeleton_th, cmap='bone_r', interpolation='none') ax2 = fig.add_subplot(1, 3, 2) imshow(skeleton_cleaned_th, cmap='bone_r', interpolation='none') ax.set_xticks([]) ax.set_yticks([]) ax2.set_xticks([]) ax2.set_yticks([]) Unfortunately I cannot share the data as it is proprietary, but I will for the next example, which is the one that does not work. There's one derivative that shows lots of detail but not continuous maxima. As a workaround I created filled contours in Matplotlib exported as an image. The image is attached. Now I want to import back the image and plot it to test: # import back image cfthdr=io.imread('filled_contour.png') # threshold using using Otsu method thresh_thdr = threshold_otsu(cfthdr) binary_thdr = cfthdr > thresh_thdr # plot it fig = plt.figure(figsize=(5, 5)) ax = fig.add_subplot(1, 1, 1) ax.set_xticks([]) ax.set_yticks([]) plt.imshow(binary_thdr, cmap='bone') plt.show() The above works without issues. Next I want to fill the white holes inside the black blobs. I thought of 2 strategies. The first would be to use opening; the second to invert the image, and then fill the holes as in here: http://scikit-image.org/docs/dev/auto_examples/plot_holes_and_peaks.html By the way, I found a similar example for opencv here http://stackoverflow.com/questions/10316057/filling-holes-inside-a-binary-o… Let's start with opening. When I try: selem = disk(1) opened_thdr = opening(binary_thdr, selem) or: selem = disk(1) opened_thdr = opening(cfthdr, selem) I get an error message like this: --------------------------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-49-edc0d01ba327> in <module>() 1 #binary_thdr=img_as_float(binary_thdr,force_copy=False) ----> 2 opened_thdr = opening(binary_thdr, selem)/255 3 4 # plot it 5 fig = plt.figure(figsize=(5, 5)) C:\...\skimage\morphology\grey.pyc in opening(image, selem, out) 160 shift_y = True if (h % 2) == 0 else False 161 --> 162 eroded = erosion(image, selem) 163 out = dilation(eroded, selem, out=out, shift_x=shift_x, shift_y=shift_y) 164 return out C:\...\skimage\morphology\grey.pyc in erosion(image, selem, out, shift_x, shift_y) 58 selem = img_as_ubyte(selem) 59 return cmorph._erode(image, selem, out=out, ---> 60 shift_x=shift_x, shift_y=shift_y) 61 62 C:\...\skimage\morphology\cmorph.pyd in skimage.morphology.cmorph._erode (skimage\morphology\cmorph.c:2658)() ValueError: Buffer has wrong number of dimensions (expected 2, got 3) --------------------------------------------------------------------------- Any idea of what is going on and how I can fix it? As for inverting (or finding the complement) and then hole filling, that would be my preferred option. However, I have not been able to invert the image. I tried numpy.invert, adapting the last example from here: http://docs.scipy.org/doc/numpy/reference/generated/numpy.invert.html I tried something like this: http://stackoverflow.com/a/16724700 and this: http://stackoverflow.com/a/2498909 But none of these methods worked. Is there a way in scikit.image to do that, and if not, do you have any suggestions? Thank you, Matteo

2 7

Finding pairs of images (homologous chromosomes)
by Jean-Patrick Pommier 15 Nov '15

15 Nov '15

Dear All, I am trying to make pairs of images from the following set of images (chromosomes sorted by size after rotation). The idea is to make a feature vector for unsupervised classification (kmeans with 19 clusters) >From each chromosome an integral image was calculated: plt.figure(figsize = (15,15)) gs1 = gridspec.GridSpec(6,8) gs1.update(wspace=0.0, hspace=0.0) # set the spacing between axes. for i in range(38): # i = i + 1 # grid spec indexes from 0 ax1 = plt.subplot(gs1[i]) plt.axis('off') ax1.set_xticklabels([]) ax1.set_yticklabels([]) ax1.set_aspect('equal') image = sk.transform.integral_image(reallysorted[i][:,:,2]) imshow(image , interpolation='nearest') Then each integral image was flatten and combined with the others: Features =[] for i in range(38): Feat = np.ndarray.flatten(sk.transform.integral_image(reallysorted[i][:,:,2])) Features.append(Feat) X = np.asarray(Features) print X.shape The X array contains *38* lines and 9718 features, which is not good. However, I trried to submit these raw features to kmeans classification with sklearn using a direct example <http://scikit-learn.org/stable/modules/neighbors.html> : from sklearn.neighbors import NearestNeighbors nbrs = NearestNeighbors(n_neighbors=*19*, algorithm='ball_tree').fit(X) distances, indices = nbrs.kneighbors(X) connection = nbrs.kneighbors_graph(X).toarray() Ploting the connection graph shows that a chromosomes is similar to more than one ... - Do you think that integral images can be used to discriminate the chromosomes pairs? - If so, how to reduce the number of features to 10~20? (to get a better discrimination) Thanks for your advices. Jean-Patrick

3 5

Re: Template matching with transparent regions
by Kevin Keraudren 12 Nov '15

12 Nov '15

Hi Stefan, Thanks for correcting me! Looking at the template matching code, I think you will have difficulties making it work with non square templates as long as you use fftconvolve(). You need to have a custom convolution code so that you can skip parts of the template as well as the corresponding parts of the image. However, what you want can be achieved in a brute force manner with a few lines of Python (as long as you are working with 2D images of reasonable size!), using: from sklearn.feature_extraction.image import extract_patches_2d from scipy.spatial import distance See example attached. Kind Regards, Kevin > On 11 Nov 2015, at 08:08, Stéfan van der Walt <stefanv(a)berkeley.edu> wrote: > > I think kwc meant that he wanted parts of the template to be ignored. I have the same problem at the moment, so will have to investigate soon if no other dev informs us how :) > > Stéfan > > On Nov 10, 2015 12:26 PM, "'Kevin Keraudren' via scikit-image" <scikit-image(a)googlegroups.com <mailto:scikit-image@googlegroups.com>> wrote: > Hi, > > What about simply masking out in the matching results the region you want to exclude: > > import numpy as np > from skimage.feature import match_template > > ... > > matching_result = match_template( image, template, pad_input=False ) > > matching_result = np.array(matching_result) > > # mask out the region you want to exclude > matching_result[mask==0] = 0 > > # coordinates of the best match > best_match = np.unravel_index( np.argmax(matching_result), matching_result.shape ) > > See this example to better understand what matching_result should look like: > http://scikit-image.org/docs/dev/auto_examples/plot_template.html <http://scikit-image.org/docs/dev/auto_examples/plot_template.html> > > Kind Regards, > > Kevin > > >> On 10 Nov 2015, at 09:11, 'kwc' via scikit-image <scikit-image(a)googlegroups.com <mailto:scikit-image@googlegroups.com>> wrote: >> >> Hi, I am new to scikit-image. I would like to use the Template Matching function in scikit-image. >> >> In my template image, there are some regions that I would like to ignore or exclude from the template matching process. I plan to change those regions to become transparent (alpha channel). Does the alpha channel have any effect on the template matching result? >> >> Thanks. >> >> >> -- >> You received this message because you are subscribed to the Google Groups "scikit-image" group. >> To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com <mailto:scikit-image+unsubscribe@googlegroups.com>. >> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>. > > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com <mailto:scikit-image+unsubscribe@googlegroups.com>. > For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>. > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com <mailto:scikit-image+unsubscribe@googlegroups.com>. > For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.

1 0

Template matching with transparent regions
by kwc 11 Nov '15

11 Nov '15

Hi, I am new to scikit-image. I would like to use the Template Matching function in scikit-image. In my template image, there are some regions that I would like to ignore or exclude from the template matching process. I plan to change those regions to become transparent (alpha channel). Does the alpha channel have any effect on the template matching result? Thanks.

3 2