scikit-image May 2017

scikit-image@python.org

15 participants
26 discussions

Whar are values returned by measure.marching_cubes

by EV Petr

Hello, I'm not familiar with scikit-image. I only want to use the function `measure.marching_cubes`, but I cannot figure out what really represent the values returned by it. `help(measure.marching_cubes)` says: *values : (V, ) array* *Gives a measure for the maximum value of the data in the local region near each vertex. This can be used by visualization tools to apply a colormap to the mesh.* What is data here (this term is too general)? I saw that values.shape[0] is the same as vertices.shape[0]. Hence each value is associated to a vertex. How is it assigned to a vertex? How are these values related to the volume or to the equation F(x,y,z)=0? (I want to extract a triangulation of an implicit surface having this equation). A second question is related to the coordinates of vertices. When I plot the trisurf defined by `vertices` and `faces` I notice that the x, y and z-range illustrated by tick values is different from the real values. More precisely: if my surface has zmin=-1 and zmax=2, on z-axis this range is [5,11]. Is there a transformation that maps the vertices coordinates, as they are returned by `measure.marching_cubes` to real coordinates on my surface? Thanks, Eve

4 years, 7 months

Re: [scikit-image] Daily wheel builds

by Stefan van der Walt

On May 26, 2017 23:38:20 Matthew Brett <matthew.brett(a)gmail.com> wrote: > I just set up daily wheel builds for scikit-image master. Thanks, Matthew! This is very helpful, especially the 32-bit builds. Best regards Stéfan

4 years, 7 months

Daily wheel builds

by Matthew Brett

Hi, I just set up daily wheel builds for scikit-image master. They run via travis-ci cron jobs, and have the dual role of testing that wheel building works, as well as testing on 32-bit Linux, and several macOS versions. The wheel builds run from the "daily" branch of the scikit-image-wheels repo: https://github.com/scikit-image/scikit-image-wheels/tree/daily See here for logs of the daily builds: https://travis-ci.org/scikit-image/scikit-image-wheels/branches The build exposed a few test failures on 32-bit - see: https://github.com/scikit-image/scikit-image/issues/2670 Cheers, Matthew

4 years, 7 months

回复： Fwd: Extract a particular area in document

by imagepy＠sina.com

Hi FWD: There are many methods to do this. as you said, you had do a transform to make sure the image in standard direction. if your image are in the standard scale, you can use a template match. but if has different scales, it will not work. but I think you can do a threshold, then binary fill, then the [no, si] will be a solid rectangle. you can do a region analysis http://scikit-image.org/docs/stable/api/skimage.measure.html#regionprops to make sure if it is your goal (by the area, perimeter and moments). I think it will be ok! you can also try the adboost, sift, and other local descriptor, and so on. but I think it is too fussy and waste to solve this problem. YXDragon ----- 原始邮件 ----- 发件人：Fabio <fdalforno(a)gmail.com> 收件人：scikit-image(a)python.org 主题：[scikit-image] Fwd: Extract a particular area in document 日期：2017年05月26日 14点42分 Hi all, my name's Fabio Dal Forno. I'm a basic scikit user and i have many question too. I've a lot of scanned document like this And i want to extract a particular area (si no, yes or no in english) see image below After that i want to train a neural networ for detect the data compiled by people, y,x coordinate for detect the means of the area (materials damage ecc ecc) and a fixed dimension image for a neural network I started to work with the following step: 1) Homography for detect document orientation after that i transform the document for rectify that 2) Homography for detect si/no area and coordinate once the first step work wery well but the second doesn't detect anything. Am i wrong reasoning ? Thanks a lot and sorry for my beginner english :) _______________________________________________ scikit-image mailing list scikit-image(a)python.org https://mail.python.org/mailman/listinfo/scikit-image

4 years, 7 months

Re: [scikit-image] Fwd: Extract a particular area in document

by Juan Nunez-Iglesias

Hi Fabio! Ach, you just gave me flashbacks to the time someone broke my car window when I was holidaying in Lido di Ostia. =P Cool work though! =) What do you mean by "using a homography" to detect the Si/No boxes? Could you provide maybe a sample image and working code? It's hard to tell what is going wrong without that context. I understand that there's privacy concerns but maybe you could fill in a blank form as test data? (Or your code to try to find the two si/no boxes that are not masked in the image below.) One thing I'll say though is that neural networks require lots and lots of ground truth data to train effectively, and is probably overkill for what you want to do. In your case, you can probably align the car images very accurately, and detecting marks outside of that should not be too hard. Good luck! Juan. On 26 May 2017, 4:42 PM +1000, Fabio <fdalforno(a)gmail.com>, wrote: > Hi all, my name's Fabio Dal Forno. > > I'm a basic scikit user and i have many question too. > > > > I've a lot of scanned document like this > > > > > > > > And i want to extract a particular area (si no, yes or no in english) see image below > > > > > > > > After that i want to train a neural networ for detect the data compiled by people, y,x coordinate for detect the means of the area (materials damage ecc ecc) > > and a fixed dimension image for a neural network > > > > I started to work with the following step: > > > > 1) Homography for detect document orientation after that i transform the document for rectify that > > 2) Homography for detect si/no area and coordinate once > > > > the first step work wery well but the second doesn't detect anything. > > > > Am i wrong reasoning ? > > > > > > Thanks a lot and sorry for my beginner english :) > > > > > > > > > > > > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image

4 years, 7 months

Fwd: Extract a particular area in document

by Fabio

Hi all, my name's Fabio Dal Forno. I'm a basic scikit user and i have many question too. I've a lot of scanned document like this And i want to extract a particular area (si no, yes or no in english) see image below After that i want to train a neural networ for detect the data compiled by people, y,x coordinate for detect the means of the area (materials damage ecc ecc) and a fixed dimension image for a neural network I started to work with the following step: 1) Homography for detect document orientation after that i transform the document for rectify that 2) Homography for detect si/no area and coordinate once the first step work wery well but the second doesn't detect anything. Am i wrong reasoning ? Thanks a lot and sorry for my beginner english :)

4 years, 7 months

Re: [scikit-image] Viewing 3D images/stacks of images

by Kesavan Subburam

from matplotlib.widgets import Slider import matplotlib.pyplot as plt import numpy as np def ddd(images): def _update_image(num): num = int(num) image = np.squeeze(images[num:num+1]) img_ax.set_data(image) fig.canvas.draw_idle() if images.ndim is not 3: raise ValueError("Not a 3D image.") Z, _, _ = images.shape fig, ax = plt.subplots() img_ax = ax.imshow(np.squeeze(images[0]), cmap="gray") ax.axis("off") sliderax = plt.axes([0.19, 0.05, 0.65, 0.03], facecolor="lightgoldenrodyellow") img_slider = Slider(sliderax, "Z", 0, Z-1, valfmt='%d', valinit=0) img_slider.on_changed(_update_image) plt.show() I'd forgotten to include the import statements. Just in case you'd like to test the code. Input image of shape (Z, N, M). On 19 May 2017 at 22:10, Kesavan Subburam <pskeshu(a)gmail.com> wrote: > Hi all, > > I'd used CollectionViewer from skimage.viewer before for viewing > microscopy Z stacks. However, I personally felt it was a bit of a pain to > use. For instance, before passing in an array of images, the image has to > be stretched, otherwise, most of the time, one would end up with a "black" > series of images. > > So, I wrote a simple function using matplotlib to view image collections, > which works quite well, and I think it can be extended to display nD images > with multiple sliders. > > def ddd(images): > def _update_image(num): > num = int(num) > image = np.squeeze(images[num:num+1]) > img_ax.set_data(image) > fig.canvas.draw_idle() > > if images.ndim is not 3: > raise ValueError("Not a 3D image.") > > Z, _, _ = images.shape > > fig, ax = plt.subplots() > img_ax = ax.imshow(np.squeeze(images[0]), cmap="gray") > ax.axis("off") > > sliderax = plt.axes([0.19, 0.05, 0.65, 0.03], > facecolor="lightgoldenrodyellow") > img_slider = Slider(sliderax, "Z", 0, Z-1, > valfmt='%d', valinit=0) > img_slider.on_changed(_update_image) > plt.show() > > > This might be a bit redundant, but I would like your thoughts on whether > this might be useful for scikit-image. Some of the plugins of > CollectionViewer can be rewritten for this matplotlib based approach in a > much more readable fashion -- for instance, the line profile plugin. > > Best, > > -- > Kesavan Subburam > -- Kesavan

4 years, 8 months

Viewing 3D images/stacks of images

by Kesavan Subburam

Hi all, I'd used CollectionViewer from skimage.viewer before for viewing microscopy Z stacks. However, I personally felt it was a bit of a pain to use. For instance, before passing in an array of images, the image has to be stretched, otherwise, most of the time, one would end up with a "black" series of images. So, I wrote a simple function using matplotlib to view image collections, which works quite well, and I think it can be extended to display nD images with multiple sliders. def ddd(images): def _update_image(num): num = int(num) image = np.squeeze(images[num:num+1]) img_ax.set_data(image) fig.canvas.draw_idle() if images.ndim is not 3: raise ValueError("Not a 3D image.") Z, _, _ = images.shape fig, ax = plt.subplots() img_ax = ax.imshow(np.squeeze(images[0]), cmap="gray") ax.axis("off") sliderax = plt.axes([0.19, 0.05, 0.65, 0.03], facecolor="lightgoldenrodyellow") img_slider = Slider(sliderax, "Z", 0, Z-1, valfmt='%d', valinit=0) img_slider.on_changed(_update_image) plt.show() This might be a bit redundant, but I would like your thoughts on whether this might be useful for scikit-image. Some of the plugins of CollectionViewer can be rewritten for this matplotlib based approach in a much more readable fashion -- for instance, the line profile plugin. Best, -- Kesavan Subburam

4 years, 8 months

build topograph network from skeleton

by imagepy＠sina.com

Hi everyone: I saw the newest version, require the networkx module. I find some function to find shortest path from image. But I want to know if there is a method to build a network from a skeleton image(2D/3D with arcs and nodes). I wrote one, but in python. it runs very slowly... yxdragon-imagepy

4 years, 8 months

回复:Re: 回复: Need help to get graph from image

by imagepy＠sina.com

I has build the networt, look tips 2. imagepy is not a function library, but a interactive framework witch can wrap numpy, scipy, scikitimage, opencv as a friendly.software. by another, please donot save picture as jpg. the dct transform may destroy the high frequent information. -------------------------------- 来自我的新浪邮箱android客户端 -------- 原始邮件 -------- 发件人:Leena Chourey 时间:2017年5月18日 13:30(星期四) 收件人:imagepy@sina.com Mailing list for scikit-image (http://scikit-image.org) 主题:Re: [scikit-image] 回复: Need help to get graph from image Thanks YXDragon for quick reply. Yes I have tried skeletonisation but hanged for further processing. hope the links u have sent will help to resolve.Wow.. great. that is exactly I need and done partially. Can u pls explain the 1.jpg--->2.jpg--->3.jpg. I want exactly the same 3.jpg (junction points). Can i call imagepy in my python script and get the o/p of it for further processing?Is it possible to convert directly image to graph data structure? On Thu, May 18, 2017 at 10:42 AM, <imagepy(a)sina.com> wrote: 1. you can do a skeleton, then use a np.ones((3,3)) to convolve. then 3 means junction, one means termin. 2 means path.2. trace from a termin. iterate all the pixel > 0. and add node if != 2! some tips:1. if you do skeleton straightly, you will get some small branch, you can remove them by graph analysis. but I have a tip, you can do a gauss blur, then threshold again. this will smooth the binary image.2. trace the path is not very simple, you need care many detail... you can look mine.3. the scikit-image's skeleton method prefers to generate branch (sometimes we want, but here we need not), you can look here. you can have a try my project imagery do some test. which can wrap scikit-image like photoshop!(but I do not know if scikit-image has a method to build the skeleton) My English is solo, forgive me. YXDragon - ImagePy--------------------------------来自我的新浪邮箱android客户端-------- 原始邮件 --------发件人:Leena Chourey时间:2017年5月18日 12:05(星期四)收件人:scikit-image@python.org主题:[scikit-image] Need help to get graph from imageDear All,I need to convert shape based image(graph_search2.jpg) into graph and then graph data structure. I have tried to filter the image and then with threshold, contour based approach access shapes(nodes), lines(edges) and arrows. but facing difficulty in getting relation b/w nodes and edges. Please help to find interest point(interest_points.jpg) and relation in node n edges.-- Thanks & Regards, -धन्यवादLeena Chourey , लीना चौरे C-DAC Mumbai सी-डैक, मुम्बई _______________________________________________scikit-image mailing listscikit-image@python.orghttps://mail.python.org/mailman/listinfo/scikit-image -------------------------------- 来自我的新浪邮箱android客户端 _______________________________________________ scikit-image mailing list scikit-image(a)python.org https://mail.python.org/mailman/listinfo/scikit-image -- Thanks & Regards, -धन्यवादLeena Chourey , लीना चौरे C-DAC Mumbai सी-डैक, मुम्बई

4 years, 8 months

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

scikit-image May 2017