EXTRACT FEATURE

Hi,
From the attached image,i need to extract(find) the features of the target object,in which the printer is the target object.
What are the target feature to be extracted ? and
how it can be extracted?
Thank you
Attachments:
photo (1).JPG ‏457 KB

Hello,
how about using SIFT descriptors and homography. You need at least 4 matched keypoints to calculate the homography.
Please see the attached example (your image). You need to install OpenCV 2.4.6 (http://opencv.org/) and add the "C:\Program Files (x86)\OpenCV 2.4.6\opencv\build\x86\vc10\bin" to system path (see here: http://windowsitpro.com/systems-management/how-can-i-add-new-folder-my-system-path).
Then just run the example. Here is a screenshot of the result.
Best regards,
K
https://decibel.ni.com/content/blogs/kl3m3n
"Kudos: Users may give one another Kudos on the forums for posts that they found particularly helpful or insightful."
Attachments:
SIFT_homography.zip ‏510 KB

Similar Messages

What are the extracted features in the IMAQ Extract Texture Feature VI?

Hi,
I am new to LabVIEW and are in need of some help. I am moving a feature extraction procedure from MATLAB into LabVIEW, this involved gathering information on textures from a gray level co-occurrence matrix (glcm). I believe that the "IMAQ Extract Texture Feature VI" will include many of the feature I want to extract but it does not seem to be listed anywhere what these features are. When I treat the entire image as one region it outputs a 1-D array of 40 values, it would be great if someone could tell me what each of these values represents or point me in the right direction.
On a side note, the original code uses a glcm that is the average of four glcm with different displacement vectors ([0 1], [-1 1], [-1 0], [-1 -1]) representing 0, 45, 90, 135 degrees. How would this be possible in LabVIEW with only one input parameter for the co-occurrence matrix?
Cheers
Note: the final results do not need to be exactly the same, but the closer the better.

Hi Ryan,
Did you find the answer to the extracted features? I am trying to do something to detect textures but I have no idea on how to use the textures VIs
Thanks
Rodrigo Cuenca
www.cidesi.com

API support for color extraction feature

Is there API support for color extraction feature? if not, is
it scheduled?

Hello, thanks for the post.
In order to provide the best user experience for color
extraction, we had to split the functionality between client-side
and server-side, so we are not able to offer an API for this
functionality at this time. We are, however, very interested in
feedback on what kinds of APIs are interesting to you and how you
are interested in using them, so folks, please let us know.
Current APIs, including the new Random theme browsing view
and to view comments to your themes, are available on the
kuler API
wiki. We've seen some really fabulous and creative use of the
kuler APIs, and we look forward to more.
Sami

Dreamweaver extract feature

Hi everybody! How are you all? In DWCC, I am using the extract feature from Photoshop which I really like and find very helpful. Just one odd thing I noticed. For example, When you load a mockup and I click on a logo. (I know I should drag it over to put in as an image but just for the sake of example) It will show a drop down of the CSS properties and I can also copy the text. I created a div called #top and I copied the properties. I then copied the text and pasted it in, in the body of the doc so it was <div id="top">Text</div> Really easy and helps with time.
I noticed, that the logo which is a name seems to not fit the width and height that I was given in the extract. It always pushes the last name to the next line. I don't know why. I find it to be very odd. I will include screen pics. When I drag the logo over as an image, there is no issue because it just comes directly from Photoshop. I have been scratching my head trying to figure out why this is doing this. If I repeat everything I just said but used the word Web Design or something else it may work but then the height might be a bit off. If I type in text for a paragraph type of thing, it seems to be fine.
I created a sample mockup just to show you what I mean.
Any idea what is going on? Shouldn't the width and height match the same as the mockup? Here are some screen pics to show you what I mean. Any help would be appreciated.
Thanks!
This first screen pic shows the extract information from the name Mark Samburg.
This second image shows what happens when I copy the styles and text. See how it bumps to the next time?
This third image shows the code.

I think there is some misinformation in this thread.
1) As far as I can tell, the reason why the OP's text breaks to the next line probably is caused by rounding the text size up instead of down. The extract screenshot is taken at 67.97% size, and when the DW screenshot is scaled down at that size, the extract and dw results look identical (minus the line break). I copied the same code, and decreased the font size by 1px; then it fits perfectly well. That is why I surmise that the font-size is rounded up instead of rounded down, and it happens to break the line in this and similar instances. This has nothing to do with Extract and Photoshop basing their calculations on a base of 72ppi, in my opinion.
2) the ppi setting for images on the web has no bearing at all on how those images are displayed in a browser, whether mobile or desktop. An image at 512px x512px with a 1000ppi setting and the same image with a 1ppi setting display identical in any browser. The save for web function in Photoshop removes this type of meta data, btw.
3) for screen work ppi has no real relevance. PPI is more confusing than helpful at this point. As far as I am concerned we should be talking about pixel pitch, which is what device manufacturers tend to use anyway, and it would prevent all the confusion in regards to dpi and other ppi related myths. Also read on below, and read up on this confusing mess here:
Pixels Per Inch Awareness and CSS Px
4) with retina screens a block of four pixels is generally treated as if it were one pixel by designers/web devs. Saying a designer must export their graphics at 72ppi or 96ppi has no meaning: the actual resolution in pixels is important. When a Photoshop mockup is setup in 960px width, the ppi itself is completely irrelevant. A header image at 960px by 400px renders on both non-retina and retina screens as exactly that in a static layout. For retina screens ideally a version at twice that resolution is available for the best quality (especially for GUI work): in this case a retina screen no longers treats a block of four pixels as one, and will use the full resolution of the 2x image to render it. (Note: the actual pixel conversion depends on the device's screen ppi, of course.)
Saying a designer should export at 96ppi to prevent pixelation on retina screens is a half-truth at best: it has little to do with ppi, but rather with the actual pixel dimensions/resolution of an image. If a device such as a Nexus 7 with a 323ppi screen displays an image meant to display at 512x512px created for regular screens, it will look quite fuzzy and unsharp. Let's say this is a 72ppi version, and now create a 96ppi version: 683x683px. That version will still result in a fuzzy looking version for retina screens.
Ideally we would have to create a 2134 by 2134px image instead to cater for a screen like that. But this PPI setting is only viable for this particular device at 323ppi. And ppi as a measuring stick quickly becomes quite redundant anyway under these circumstances.
Such large imagery is generally avoided, because visually most of us would not be able to discern the difference between this and a two-up image on small screens. Bandwidth also plays an important role, of course, with such a large version congesting mobile traffic too much.
Thus, we tend to choose the middle-ground for photos: 1024px by 1024px (around 150ppi, give or take). A factor of two. For GUI work or very sharp illustrative artwork this may not work very well, however. In those cases a higher resolution version would become a requirement.
That is why anyone interested in exporting bitmap work from Photoshop should be using Generator: it makes it very simple to export a low res and 2x version of the artwork (and other versions with varying formats and sizes, if need be). Then we would use an image picturefill to decide which version to display depending on the screen/device type.
When fluid design come into play, a designer should create the non-retina version at the highest resolution that is required for the maximum width. In code this can then be set to a (max-)width of 100%, so that it accommodates the layout width automatically. For retina screens a 2x version again would work well.
Another option is to save a giant version at a very low jpg quality, which can work quite well, depending on the photo. And I have failed to mention the type of image formet in relation to the best compression algorithm - which would take as much text to write about ;-)
In a nutshell, over-generalizing the web export procedure without more quantification and qualification is to be avoided. For a certain category of graphics SVG would be the best option. For others still png 8 or 24 bit. A nice comparison writeup:
retina-revolutie-follow-up
Unfortunately in Dutch, but still quite interesting.
5) the save for web function is indeed in dire need of an update, that much is true. According to W3C standards, CSS2/3 work on the basis of a 96ppi standard, and SfW is still stuck in CSS1 times. On the other hand, ppi does nothing for images for the web, so... I would not say this is much of an issue in real practical sense, as I have explained above.
More importantly, it still does not even support 8bit png files with full transparency(!). This is unforgivable. Really, it is. Another reason to switch to Generator, which does support that option. But honestly, even Generator cannot compete with the likes of Color Quantizer for png optimization, and RIOT (Radical Image Optimization Tool) for jpg optimization. Avoid Save For Web for meaningful optimization, and instead either save out a full quality png file for later optimization in CQ or RIOT, or use Generator, and optimize those a second time in CQ or RIOT. Or other tools such as ImageOptim, PngCrush, and so on. Anything BUT SfW, which is sadly responsible for huge energy wastage across the net, because most designers have no inkling how bad it is in terms of image optimization.
Images extracted with that extract function should also be checked again for optimization, by the way.

Dreamweaver extract feature is broken

The Dreamweaver CC 2014 extract feature is not functional. Please Help!
I have renamed folders based on other posts, but nothing seems to work.
This cloud is broken or holding too much water:) I have been able to get Photoshop files to appear in a separate browser — different cloud.

Hi Mark,
Please post a snapshot of error for better understanding of the issue.
What did you mean by cloud is broken? and Photoshop files appear in separate browser?
Please elaborate more on it.
Thanks,
Lalita

The "Extract" feature is not showing up anywhere in my Dreamweaver CC 2014, why ?

Impossible to have the feature anywhere in my Dreamweaver and I have the latest of the latest version !
Tried everywhere in the workspace, even in "window" it doesn't show up !
Even in "edit>preferences" I don't have a line for "Extract" ! Why ? And the tutorial when launching shows me that I am supposed to be able to use it...
Thanks

Thank you for taking the time to answer me, but as I said clearly in my post : "Tried everywhere in the workspace, even in "window" it doesn't show up !"
See attached screenshot :

How do I use Flickr images in the Color Extraction feature? How do I browse?

From the camera mode, tap on the image icon in the lower left corner of the app. From there, select Flickr in the options menu. You can use the search box to search Flickr for the type of image you are looking for. Browse the results and tap the image to download it to be used for color extraction.

The Java Advanced Imaging API would seem to have some classes for working with .tiff images.
http://java.sun.com/products/java-media/jai/forDevelopers/jai-apidocs/

Where is the extract filter feature?

I used the extract feature frequently with CS2. Then I upgraded to CS5,
and even though the help menu said it is to be found in the filter drop-down
menu, it isn't there. Must I extract with CS2, save it, and then reopen it with CS5?
That certainly seems like a step backward.

Yes, but the Help Files also said this at the very top of that page:
To more effectively extract
objects from backgrounds, use the Refine Edge option for selections.
(See Refine
selection edges.) If you prefer the optional, Extract plug-in
described
below, download the Windows-only plug-in here.

Extract Texture feature output

I'm trying to use the IMAQ Extract Texture Feature VI. In reading the help file I see that the Texture Feature output "is a 2D array that contains the extracted feature. The number of rows in the array corresponds to the number of windows used in the image." What this doesn't explain, is what are the columns of the array? I used all the default inputs and got an output with the expected number of rows and 40 columns. Can anyone point me to documentation of some kind that can describe what's in these columns?
Thanks!
Teresa
Solved!
Go to Solution.

Ok, I was able to answer my own question. If you look on the NI website there is a special page for "NI Vision Concepts Help" http://digital.ni.com/manuals.nsf/websearch/AA0ADE1B25EC1CD086257B1F00598BB6 . If you download the zip file on this page there's a description of the texture feature extraction algorithm that says:
"The texture defect detection algorithm extracts five Haralick features—entropy, dissimilarity, contrast, homogeneity, and correlation—from the GLCM calculated at each partition of the subband texture."
"The number of feature vector elements extracted to represent a texture sample is equivalent to the number of selected wavelet subbands multiplied by 5 (the number of Haralick features extracted from each wavelet subband). For example, if an application uses all 8 subbands, the size of the resulting feature vector is 40. "
I'm annoyed that this was so hard to find though. Buried in a zip file? Really NI?

ABAP Query Extracts

hello experts ,
i am using ABAP query (sq01) infoset ( logical database ).Is it possible to add extract feature in the abap query selection screen . if possible then how?
thanks
vivek

Hi Vivek,
Its not possible. Whenever you activate your query it generates a standard program & its not modifiable.
Regards
Abhii

The option to extract pages from a PDF document as described does not appear for me.

Im currently running Acrobat Pro XL and the option to extract pages from a PDF document as described in the below tutorial does not appear for me. Please help!
Extracting pages from a PDF
https://acrobatusers.com/tutorials/extracting-pages

Typically if the extract feature is not present then the application is not Acrobat Pro.
Be well...

My feedback on Acrobat XI new extract capabilities

I'm specialized in winter sport equipment review (journalist) and nowadays I get about every datas in pdf catalogs. I extract every specs, technos and texts (100+ brand catalogs a year).
My present Acrobat is pro 9 and extract feature is rarely ok. So I use side solutions like cogniview or able2extract, and also online solutions like abbyoonline, onlineocr.
I was very very interesed in version XI because of the export new fonctionality. So I give acrobat XI a try. Acrobat XI is not bad, but not enough for me. Features are too simple. You draw a square around datas, then there is quite no more options, so you get what you get ! I tried it on 10 tables to xls and I get only 4 good results.
Many table layouts have been rearanged in orientations, type sizes, new lines, background colors. With cogniview and/or able2extract, you can go further manually, Your redraw data structure with lines, colomns and at the end you get it ! It's not perfect because you can't define the exact symbol set you want to use. Should be so cool to draw a square around something and to say "when you see that write this !". A far far dream today...
So my Acrobat 9 is still ok I'll wait for a Accrobat XII. Ready to buy it if me wants are in !

just sign in and download your software, https://creative.adobe.com/
you should actually do that on a second computer, too. you will be given the option to sign out of all your other cc accounts so the thief will use their use of your adobe cc programs at that point.

Extract pages in same PDF version as original document

That's it!
It is annoying to extract pages from a PDF 1.3 doc. and have the resulting files be in PDF 1.6.
It is also somewhat strange that the format is not even the native 1.7 format from Acrobat 8.1.1, which I'm using...
The most useful implementation would be to have a choice of output format with the original doc. format set as default.
I would also like a keyboard shorcut for the Extract feature, as requested in other posts...
/eac

Optimal application behaviour is subjective based on the user's needs.
My request stems from needing to split PDF documents into single-page documents without altering the format.
Adding a preference setting for the default behaviour would allow all users to have it their way: options to preserve format or save as specific format.
In addition, the save dialogue should contain an option to select the desired format, with the default being set from the preference setting.
/eac

Plugin for renaming extracted pages?

Does anyone know of a plugin I can use to rename pages extracted from a PDF? I work on long magazines and an occasional catalog and have to submit single-page PDFs to my printing service. I usually make one giant PDF and then extract all the pages in Acrobat. The only problem is that Acrobat names them
filename 1.pdf
filename 2.pdf
filename 3.pdf
etc. Whereas my printing service wants the filenames to adhere to a rigid format of page number first and no spaces. Example:
001filename.pdf
002filename.pdf
I definitely don't have time to manually rename them, and I believe we get charged a fee since I don't follow that format. Just curious if there are any plugins to change the default naming in Acrobat's extract feature.
Thanks, Phyllis

Just in case someone else needs this information, I've found the answer. :-) You don't need an extra utility -- you can do the renaming with Adobe Bridge. I haven't really used Bridge that much, but it's got a great batch renaming feature. So I just extract them with Acrobat and then rename with Bridge.....
Phyllis

Extracting original audio from .avi movie

I imported an .avi movie. The original sound plays, but does not show as a green strip. There is no way to delete/extract it. I called Apple Tech Support and they suggested, I buy "Final Cut" at $1000 a pop. Does anyone have a better idea?

As I originally stated, I tried everything, including calling Apple Tech Support. They said: "It cannot be done with iMovie 2009, you need to buy First Cut."
I found a way to remove the embedded audio with iTunes Professional [upgrade]. A friend told me Pinacle Studio [PC] will also do it. However, I am very unhappy that Apple removed the embedded audio extraction feature when they went from iMovie HD to iMovie 2009.
I think they want to sell more First Cut Software. I think they are making a mistake. They will **** people off.

EXTRACT FEATURE

Similar Messages

Maybe you are looking for