I have a Postgres database containing photographs and I want to let people search them by colour. I already have multiple colours (1-5) defined for each photograph, and I'm using LAB colours (perceptual colour space, defined in three dimensions: lightness plus two colour dimensions).
My question is: what is the best way to do this in Postgres? It's essentially a three-dimensional search, so should I use a spatial index?
My requirements are:
run a bounding box search by colour (find photos with colours within distance X of colour Y).
rank results by distance from colour X (return photos with colours closest to colour Y first)
r-tree-like performance.
I've built a proof-of-concept using an rtree index in Python, and it's working pretty well. I'm just not sure how to replicate it using Postgres tables.
I see a couple of options. PostGIS has multi-dimensional geometry types which may give exactly what you are looking for. You could do bounding boxes over 3d points, for example. That would be easiest.
PostGIS is a Geospacial add-on but it can be used for a large number of other things too. What you are looking at doing is spacial even if not GIS and this may be the best tool for the job. PostGIS would handle all your indexing needs also including distance searches and the like.
If this is not acceptable you could write your own types. You'd probably want to write GiST operators as well for them and there is a fairly large learning curve there.
Related
I need to train R-CNN on my dataset. Above Image is an example in which first column contain path to that image and second column contain coordinates of bounded box(ROI). How to get those coordinates in matlab. As my dataset is large so how those coordinates can be extracted by pointing manually.
for example if i am training R-CNN foe stop signs then second column contain coordinates of bounded box containing stop sign in whole image.
I do not know which version of MATLAB you are running, but I'm assuming it is fairly new (R2017a and later). Also, by 'how to get the coordinates', I assume you mean 'how to determine' or 'how to assign' the coordinates.
I believe what you need to do is to use one of the image labeling Apps called
imageLabeler
to annotate rectangles in your training images. You either do this manually if that's amenable, or you need to use automation algorithms if you already have a detector that does something similar. See this page for more details:
https://www.mathworks.com/help/vision/ug/create-and-import-an-automation-algorithm-for-ground-truth-labeling.html
Once you have the results of labeling stored in a groundTruth object, you would need to use something like objectDetectorTrainingData to create the table you are looking for.
See https://www.mathworks.com/help/vision/ug/train-an-object-detector-from-ground-truth-data.html for more details.
I was visualizing my vorticity vector field and notice that I am not able to see the pattern without zooming in as there are too many glyphs and are too packed.
Currently, I am using a calculator to combine X,Y,Z vorticity field into a single vector field using the calculator. Take a slice of it and do a glyph filter visualizing all points on the plane.
I notice that one possible way is to visualize a curved glyphs and scale up a little bit to make it more noticeable, but not sure how to do that. Does anyone know whats the steps to do that? Or any other suggestions?
TIA
Have you tried reducing the Maximum Number of Sample Points property in the Properties Panel when the Glyph filter is selected in the Pipeline Browser? You may also want to change the Scale Factor property to change the length of the glyphs.
For my application, I want to interpolate between two images(CT to PET).
Therefore I map between them like that:
[X,Y,Z] = ndgrid(linspace(1,size(imagedata_ct,1),size_pet(1)),...
linspace(1,size(imagedata_ct,2),size_pet(2)),...
linspace(1,size(imagedata_ct,3),size_pet(3)));
new_imageData_CT=interp3(imagedata_ct,X,Y,Z,'nearest',-1024);
The size of my new image new_imageData_CT is similar to PET image. The problem is that data of my new image is not correct scaled. So it is compressed. I think the reason for that is that the pixelsize between the two images is different and not involved to the interpolation. So for example :
CT image size : 512x512x1027
CT voxel size[mm] : 1.5x1.5x0.6
PET image size : 192x126x128
PET voxel size[mm] : 2.6x2.6x3.12
So how could I take care about the voxel size regarding to the interpolation?
You need to perform a matching in the patient coordinate system, but there is more to consider than just the resolution and the voxel size. You need to synchronize the positions (and maybe the orientations also, but this is unlikely) of the two volumes.
You may find this thread helpful to find out which DICOM Tags describe the volume and how to calculate transformation matrices to use for transforming between the patient (x, y, z in millimeters) and volume (x, y, z in column, row, slice number).
You have to make sure that the volume positions are comparable as the positions of the slices in the CT and PET do not necsesarily refer to the same origin. The easy way to do this is to compare the DICOM attribute Frame Of Reference UID (0020,0052) of the CT and PET slices. For all slices that share the same Frame Of Reference UID, the position of the slice in the DICOM header refers to the same origin.
If the datasets do not contain this tag, it is going to be much more difficult, unless you just put it as an assumption. There are methods to deduce the matching slices of two different volumes from the contents of the pixel data referred to as "registration" but this is a science of its own. See the link from Hugues Fontenelle.
BTW: In your example, you are not going to find a matching voxel in both volumes for each position as the volumes have different size. E.g. for the x-direction:
CT: 512 * 1.5 = 768 millimeters
PET: 192 * 2.6 = 499 millimeters
I'll let to someone else answering the question, but I think that you're asking the wrong one. I lack context of course, but at first glance Matlab isn't the right tool for the job.
Have a look at ITK (C++ library with python wrappers), and the "Multi-modal 3D image registration" article.
Try 3DSlicer (it has a GUI for the previous tool)
Try FreeSurfer (similar, focused on brain scans)
After you've done that registration step, you could export the resulting images (now of identical size and spacing), and continue with your interpolation in Matlab if you wish (or with the same tools).
There is a toolbox in slicer called PETCTFUSION which aligns the PET scan to the CT image.
you can install it in slicer new version.
In the module's Display panel shown below, options to select a colorizing scheme for the PET dataset are provided:
Grey will provide white to black colorization, with black indicating the highest count values.
Heat will provide a warm color scale, with Dark red lowest, and white the highest count values.
Spectrum will provide a warm color scale that goes cooler (dark blue) on the low-count end to white at the highest.
This panel also provides a means to adjust the window and level of both PET and CT volumes.
I normally use the resampleinplace tool after the registration. you can find it in the package: registration and then, resample image.
Look at the screensht here:
If you would like to know more about the PETCTFUSION, there is a link below:
https://www.slicer.org/wiki/Modules:PETCTFusion-Documentation-3.6
Since slicer is compatible with python, you can use the python interactor to run your own code too.
And let me know if you face any problem
I have 42 variables and I have calculated the correlation matrix for them in Matlab. Now I would like to visualize it with a schemaball. Does anyone have any suggestions / experiences how this could be done in Matlab? The following pictures will explain my point better:
In the pictures each parabola between variables would mean the strength of correlation between them. The thicker the line is, the more correlation. I prefer the style of picture 1 more than the style in picture 2 where I have used different colors to highlight the strength of correlation.
Kinda finished I guess.. code can be found here at github.
Documentation is included in the file.
The yellow/magenta color (for positive/negative correlation) is configurable, as well as the fontsize of the labels and the angles at which the labels are plotted, so you can get fancy if you want and not distribute them evenly along the perimeter/group some/...
If you want to actually print these graphs or use them outside matlab, I suggest using vector formats (eg eps). It's also annoying that the text resizes when you zoom in/out, but I don't know of any way to fix that without hacking the zoom function :/
schemaball % demo
schemaball(arrayfun(#num2str,1:10,'uni',false), rand(10).^8,11,[0.1587 0.8750],[0.8333 1],2*pi*sin(linspace(0,pi/2-pi/20,10)))
schemaball(arrayfun(#num2str,1:50,'uni',false), rand(50).^50,9)
I finished and submitted my version to the FEX: schemaball and will update the link asap.
There are a some differences with Gunther Struyf's contribution:
You can return the handles to the graphic object for full manual customization
Labels are oriented to allow maximum left-to-rigth readability
The figure stretches to fit labels in, leaving the axes unchanged
Syntax requires only correlations matrix (but allows optional inputs)
Optimized for performance.
Follow examples of demo, custom labels and creative customization.
Note: the first figure was exported with saveas(), all others with export_fig.
schemaball
x = rand(10).^3;
x(:,3) = 1.3*mean(x,2);
schemaball(x, {'Hi','how','is','your','day?', 'Do','you','like','schemaballs?','NO!!'})
h = schemaball;
set(h.l(~isnan(h.l)), 'LineWidth',1.2)
set(h.s, 'MarkerEdgeColor','red','LineWidth',2,'SizeData',100)
set(h.t, 'EdgeColor','white','LineWidth',1)
The default colormap:
To improve on screen rendering you can launch MATLAB with the experimental -hgVersion 2 switch which produces anti/aliased graphics by default now (source: HG2 update | Undocumented Matlab). However, if you try to save the figure, the file will have the usual old anti-aliased rendering, so here's a printscreen image of Gunther's schemaball:
Important update:
You can do this in Matlab now with the FileExchange submission:
http://www.mathworks.com/matlabcentral/fileexchange/48576-circulargraph
There is an exmample by Matlab in here:
http://uk.mathworks.com/examples/matlab/3859-circular-graph-examples
Which gives this kind of beautiful plots:
Coincidentally, Cleve Moler (MathWorks Chief Mathematician) showed an example of just this sort of plot on his most recent blog post (not nearly as beautiful as the ones in your example, and the connecting lines are straight rather than parabolic, but it looks functional). Unfortunately he didn't include the code directly, but if you leave him a comment on the post he's usually very willing to share things.
What might be even nicer for you is that he also applies (and this time includes) code to permute the rows/columns of the array in order to maximize the spatial proximity of highly connected nodes, rather than randomly ordering them around the circumference. You end up with a 'crescent'-shaped envelope of connecting lines, with the thick bit of the crescent representing the most highly connected nodes.
Unfortunately however, I suspect that if you need to enhance his code to get the very narrow, high-resolution lines in your example plots, then MATLAB's currently non-anti-aliased graphics aren't quite up to it yet.
I've recently been experimenting with MATLAB data and the D3 visualization library for similar graphs - there are several related types of circular visualizations you may be interested in and many of them are interactive. Another helpful, well-baked, and freely available option is Circos which is probably responsible for most of the prettier versions of these graphs you've seen in popular press.
I would like to check whether an image has a lot of homogeneous areas. Therefore I would like to get some kind of value of an image that declares a ratio for images depending on the amount/size of homogeneous areas (e.g. that value could have a range from 0 to 5).
Instead of a value there could be some kind of classification as well.
[many homogeneous areas -> value/class 5 ; few homogeneous areas -> value/class 0]
I would like to do that in perl. Is there a package/function or something like that?
What you want seems to be an area of image processing research which I am not familiar with. However, GraphicsMagick's mogrify utility has a -segment option:
Use -segment to segment an image by analyzing the histograms of the color components and identifying units that are homogeneous with the fuzzy c-means technique. The scale-space filter analyzes the histograms of the three color components of the image and identifies a set of classes. The extents of each class is used to coarsely segment the image with thresholding. The color associated with each class is determined by the mean color of all pixels within the extents of a particular class. Finally, any unclassified pixels are assigned to the closest class with the fuzzy c-means technique.
I don't know if this is any use to you. You might have to hit the library on this one, and read some research. You do have access to this through PerlMagick as well. However, it does not look like it gives access to the internals, but just produces an image based on parameters.
In my tests (without really understanding what the parameters do), photos turned entirely black, whereas PNG images with large areas of similar colors were reduced to a sort of an average color. Whether you can use that fact to develop a measure is an open question I am not going to investigate ;-)