Comparing images using vector

Sir and Maam! I just want to ask something.
my project is all about face recognition. i got the ideas from here how to make it. well i am on the last part of it now.
well what i do is this based on my research on forums.
1. RGB to YUV
2. get the location of the face
3. grayscaled
4. cut the location of the face.
5. and i discrete cosine transform the cut off image.
so i am on the final part. they say that it is easy to know if the face are the same using vector. but i cant get it.
can you guys see if the operation is correct.
i will attatch the image of my vi for subtracting the two vectors.
Attachments:
w.png ‏7 KB

Hello there,
I am having trouble following along with what you are trying to ask. Also, keep in mind I am not aware of what research you have done on the forums or what sources you have been referring to.
With that let me start by asking, when you say you are at the final part, would that be the 6th step? If yes, am I to understand correctly that you have already recognized the face, cut it out, and preformed a discrete cosine transform of the cut out?
I have looked at the image you attached for subtracting the two vectors, and assume that this is the last step you are referring to. Please provide more information on what it is you are trying to do for this final step as I cannot say if the operation is correct if I do not understand what it is you are trying to do, and I do not want to assume.
Matthew R.
Applications Engineer
National Instruments

Similar Messages

Organizer 12 - how to zoom over 100%, compare images, use the Magic Trackpad?

English Version below.
Hallo zusammen. Ich bin von Apple Aperture auf PSE 12 gewechselt. Nun habe ich meine 11500 Fotos im Organizer und ein paar Startschwierigkeiten.
Ich habe in den Voreinstellungen angegeben, dass man Fotos auf über 100% skalieren darf. Das funktioniert aber nicht. Fotos werden immer nur im Fenster eingepasst, wenn der Zoomhebel ganz nach rechts steht.
Einzig in der Vollbildansicht kann man über 100% skalieren. Dies funktioniert aber auch, wenn ich die Voreinstellung dazu nicht gesetzt habe.
ie vergleiche ich mehrere Bilder? Im Handbuch steht beschrieben, dass man dazu aus dem Menü Ansicht -> Bilder Vergleichen wählen soll. Diesen Menüpunkt gibt es bei mir nicht.
Auch bei dem Druck auf F12 geschieht bei mir nichts. Kann ich die Tastenfunktionen irgendwo neu zuweisen?
Zu guter Letzt: kann ich die Zoom & Pinch Funktion des Magic Trackpads irgendwie einschalten? Das vermisse ich sehr. Da war Aperture sehr Bedienerfreundlich.
English Version
Hi there. I switched from Apple Aperture to PSE 12. Now I have my 11500 photos in the Organizer and a few teething problems.
I have specified in the preferences that you can resize photos to over 100%. This does not work. Photos are always fitted only in the window when the zoom lever is all the way to the right.
Only in the full-screen view you can scale over 100%. However, this also works if I have not set the setting in the preferences.
How do I compare multiple images? The manual describes to select from the View menu -> Compare images. This menu item does not exist for me. (German Version)
Also by pressing F12, nothing happens. Can I re-assign the button functions somewhere?
Last but not least: How can I use the zoom & pinch from the Magic Trackpad? I 'm missing that very much. Aperture was very user friendly with that.

Hello Barbara and thank you for your answer.
I've ment side by side view. And it works by pressing F12 + cmd. Thanks.
Zoom & Pinch only works in the Editor, but not very well or smooth. And I couldn't tell you how often the Editior rotate my images unwanted ...
In the preferences of the Organizer, there is no point for a scroll wheel. And I'm really not able to zoom in the organizer over the "fitts in screen" size.
And I've found some more problems Sometime I'd like to export my images without any Exif/Metadata Information and resized. I haven't found any point do do so.
At the Moment, I'm thinking about switching back to Aperture, where I came from - I've tried to use PSE 'cause of Adobes Raw Converter, which is much much better with Olympus RAW Files than Apertures.

Organizer 12 - how to: zoom over 100%, use the Trackpad, compare images?

English Version below.
Hallo zusammen. Ich bin von Apple Aperture auf PSE 12 gewechselt. Nun habe ich meine 11500 Fotos im Organizer und ein paar Startschwierigkeiten.
Ich habe in den Voreinstellungen angegeben, dass man Fotos auf über 100% skalieren darf. Das funktioniert aber nicht. Fotos werden immer nur im Fenster eingepasst, wenn der Zoomhebel ganz nach rechts steht.
Einzig in der Vollbildansicht kann man über 100% skalieren. Dies funktioniert aber auch, wenn ich die Voreinstellung dazu nicht gesetzt habe.
ie vergleiche ich mehrere Bilder? Im Handbuch steht beschrieben, dass man dazu aus dem Menü Ansicht -> Bilder Vergleichen wählen soll. Diesen Menüpunkt gibt es bei mir nicht.
Auch bei dem Druck auf F12 geschieht bei mir nichts. Kann ich die Tastenfunktionen irgendwo neu zuweisen?
Zu guter Letzt: kann ich die Zoom & Pinch Funktion des Magic Trackpads irgendwie einschalten? Das vermisse ich sehr. Da war Aperture sehr Bedienerfreundlich.
English Version
Hi there. I switched from Apple Aperture to PSE 12. Now I have my 11500 photos in the Organizer and a few teething problems.
I have specified in the preferences that you can resize photos to over 100%. This does not work. Photos are always fitted only in the window when the zoom lever is all the way to the right.
Only in the full-screen view you can scale over 100%. However, this also works if I have not set the setting in the preferences.
How do I compare multiple images? The manual describes to select from the View menu -> Compare images. This menu item does not exist for me. (German Version)
Also by pressing F12, nothing happens. Can I re-assign the button functions somewhere?
Last but not least: How can I use the zoom & pinch from the Magic Trackpad? I 'm missing that very much. Aperture was very user friendly with that.

You cannot zoom in on an more that fitting the screen space will looking at images in the Organizer's image well. This size may be much smaller than 100%.
There is no preference to "resize photos to over 100%".
There is also not a Compare option in the View menu.
To do either of these options (from the Organizer) you have to be in the Full Screen mode. From here, you can use the mouse wheel (or trackpad) to zoom up to 1600%. You can also compare images by selecting the Side-by-Side or Top/Bottom display from the "View" Options button on the Full Screen options bar.
To use the F-keys, you have to add the Command key, so Command+F12 will open the Full Screen Mode in side-by-side compare view.
If you purchased Photoshop Elements as a replacement for Aperture, there are a lot of differences that you will have to adjust to. Photoshop Elements and Aperture are hardly complementary applications. If you were looking for something like Aperture, you should try Lightroom instead.

Please help, cannot compare images in full screen mode

I am not able to compare 2 images in full screen mode. I am using a MBP and a ACD 23". The main image will remain in the ACD while the image being compare is being shown in the MBP.
I am able to compare images in the viewer mode.
My main display is the ACD 23 which I selected using the display preferences under 'Arrangement' with the menu bar located on the ACD. In Aperture, Secondary Viewer is set to Blank.
I had also tried different combinations of Main Viewer and Secondary Viewer settings but was not successful.
Am I missing something? Anyone have the solution?Thanks in advance for any help.
norheng

Anyone have any suggestions to using Aperture to edit images using a MBP and ACD so that it best utilize Aperture full screen mode in comparing and narrowing down to the pick.
I am now trying to use the MBP as the main display and ACD as the secondary. This way I am able to compare images with the MBP in the Browser mode and the ACD display the compare images full screen. It work with the secondary display - ACD set to mirror, span and alternate mode.
It would be best if only one screen is use and hence save energy consumption (reducing global warming and bill, if it help).
Thanks in advance for any suggestions or and solutions.
norheng

[VB, CS3] How to place images using wildcards in filenames

Dear all,
Can anyone think of a way to place an image using a wildcard in the filename?
I can place the image fine if the complete filename is known, but what if the filename changes from day to day? - eg it could be House1234.tif for one document then House56r.tif for another. So I'd like to place House*.tif sort of thing. The files are in different folders, of which the path is known.
Any ideas?
Nigel

Get the contents of the whole folder in an array, then you can check each item to see if it matches some criterium. In Javascript (VB should be similar):
function find_file (dir, mask)
   var f = Folder (dir).getFiles ('*.*');
   for (var i = 0; i < f.length; i++)
   if ( // compare f[i].name and mask here
      return f[i];
This returns the first file in "dir" that matches "mask".
Peter

Comparing images from different cameras

I recently tried to compare images taken with a Nikon D300 (12.3 mgapixel-small sensor), D800 (36 megapixels-large sesnsor) and D4 (16.1 megapixels - large sensor) using the LR4.1 Compare function with not very satisfactory results (iMac 27, Lion). I took 4 images each of the same subject with a 60mm f2.8 Nikon lens on a tripod at F8, at 200,400,800 and 1600 ISO and imported the images into LR. Selecting images of the same ISO and using the Compare function, they all looked pretty much the same so I viewed them at 1:1 and 2:1 trying to get a closer look at the fine details to check for sharpness and noise. Regardless of how I cropped the images, I couldn't get any of them from the different cameras to be the same size on the screen in the closeup views. I presume the combination of sensor size and megapixel capture on the sensors caused this problem but I don't have the technical expertise to either understand (or care). Is there a way in LR to overcome these differences? If so, how? Thanks for any help.

Obviously, Lightroom was not designed to handle this specialized situation. This comparison would be a piece of cake in Photoshop (or Photoshop Elements or most any other true editor), but I cannot think of a way to make this happen in Lightroom.

Bitmap images to Vectored images

Is there a way really one can use to convert bitmap images into vectored images?
I ask this question in response to a problem that I have. I take photos using a camera of resolution 72 pixels per inch and then process them in a program I have written in Java. I then print cards as the output of the system processing. The picture quality I then produce is a worry. The images are blurred and they rather look like zoomed bitmap images. Its not very disgusting but its something the clients are complaining about.
I have tried editing the images in different photo editors but it still remains a question. I have tried to change the resolution of the images but still the problem persists. This the led me to thinking that the reasons could be:
1. the printer I am using is printing at a different resolution with the resolution of the camera.
2. the images are bitmaps and so since I am changing the height and width (propotionally though), that could be leading to distortion of the images slightly. Hence I am looking for a bitmap to vector image convetor.
Suggestions with regards to where I could be going wrong are very much welcome.
Thanx in advance
kernan

Is there a way really one can use to convert bitmap
images into vectored images?
I ask this question in response to a problem that I
have. I take photos using a camera of resolution 72
pixels per inch and then process them in a program I
have written in Java. I then print cards as the
output of the system processing. The picture quality
I then produce is a worry. The images are blurred and
they rather look like zoomed bitmap images. Its not
very disgusting but its something the clients are
complaining about. And what makes you think that vectorizing an image is actually an improvement?
Suggestions with regards to where I could be going
wrong are very much welcome.Get a better camera.

Extract images using itext library (Urgent)

Hi All,
Please send some sample example of extracting embedded images using iText library.
Thank in advance.
- Ganesh

@Avinash – I'd use InDesign (the re-opened IDML file) and some ExtendScript scripting for that task.
You have to distinguish between three basic data types:
1. Pixel only
2. Vector only
3. Mixed pixels and vector
Pixel only: could stem from a variety of source files: from PSD, TIFF, JPEG, PNG, EPS…
Vector only: EPS, WMF, PDF
Mixed ones: EPS, PDF, PhotoShop (//EDIT: Not to forget PSD!!)
It depends on your InDesign version which type of file you can write (export):
PDF, EPS, JPEG, PNG*
So, for an unclear situation (the nature of the placed file) I always would chose PDF as export format with transparency intact: PDF/X-4.
And proceed on with the exported PDFs. You could parse them to PhotoShop writing pure pixel formats if you want.
You could also look for Adobe Illustrator meta data and try to open it in Adobe Illustrator.
Here an example for what can be done with ExtendScript and pasted or embedded images by Peter Kahrel:
Peter Kahrel
Write pasted and embedded images to disk
http://www.kahrel.plus.com/indesign/unembed_images.html
However, this script is missing PDFs as export format.
Layered embedded PhotoShop or Adobe Illustrator files would also be a very special case that should be treated separately.
Color Management in these cases could be a problem too.
From pure IDML, I see no chance…
(but maybe someone else)
Uwe
Message was edited by: Laubender

Transform implementation goes wrong when I try transforming a image through script and save the new image using a HTML to canvas plugin.

Transform implementation goes wrong when I try transforming a image through script and save the new image using a HTML to canvas plugin. The rotation comes up fine, but the origin of the transformation is faulty (compared to other browsers like Chrome & IE)

A good place to ask advice about web development is at the mozillaZine "Web Development/Standards Evangelism" forum.
*http://forums.mozillazine.org/viewforum.php?f=25
The helpers at that forum are more knowledgeable about web development issues.
You need to register at the mozillaZine forum site in order to post at that forum.

Help converting raster product image to vector

I have an image of one of our products that I need to convert to a vector image for the use in a video. I'm not too skilled in IA and everytime I try to use Live Trace (w/ various settings) the results are not ideal. I've included a jpeg of the product. I have a hi-res raw file to work with. Can anyone provide me with advice on the best way to go about converting the image to vector? Thank you in advance!
Mel

Sounds like a poorly defined project then. You need to nail down the client's requirements and expectations.
If the requirements are simply for a scaleable 2D drawing, then a number of rendering methods will work. But if the animation envisioned calls for the cylindrical object to rotate (about any axis other than the line of sight), then doing that kind of frame-by-frame artwork in Illustrator will be inefficiently tedious.
Illustrator's 3D Effect can extrude the semicircle or split circle section. And creating a Blend between two copies of a 3D Effect object with different orientations can achieve the series of frame-by-frame images. But it cannot do the notch as one extruded object. Nor can it extrude two separate paths in the same 3D coordinate system at different positions along the extrusion axis. Nor can it revolve two paths at different degrees in the same 3D space. Plus, its rendering of reflective metalic surfaces is not as versatile as would be required, so you'd have to resort to artwork maping, which would not likely be very convincing on a rotating object.
So if the tube-like object is to spin, Illustrator is a poor choice of a program to use for it.
JET

Which is the best way to compare images?

hi all,
My question is that if i want to compare two images in midp2.0 and obtain perceptual diff. between those two images,then what should i do?where to start with?pixel-by-pixel comparison or any othere approach i should use?please guide me through this.

hi gnat,
finaly i have thought a approach just tell me that how efficient it will be,
Say We Have To images,among which first is a base image and the other image is an image to which we are comparing our base image.
1.first of all i am gointg to obtain Red,green,blue And alpha values for both image using getRGB() function and then store those values in integer arrays.
2.after doing this i am going to buil a compar() function,in which
say,r1 is the array of red values for all pixels in image1(baseimage) and r2 is the array of red values for all pixels in image2.
so what i m going to do in compare() function is substract each values of r1 from each values of r2,(r1-r2).
and assign a percentage based on the difference (r1-r2).and do the same for green values and blue values.so is it ok or not?

Priority Queue using Vector

I am trying to develop Priority Queue using vector. For inserting elements I need to check if the inserted objects have greater key I want to shift them. In short, I want to store objects in vector in sorted manner (sort by key). I need help for that.
The code I have created so far is like this:
import PriorityQueue;
import java.util.*;
import java.io.*;
public class PQItem
     private Object e;
     private int k;
     public PQItem(int key, Object element)
          k=key;
          e=element;
     public int key(){ return k;}
     public Object elelment(){ return e;}
     public void setkey(int key){k=key;}
     public void setElement(Object element){e=element;}
     public static void main(String args[])
          FileReader fr;
          fr = openfile();
     public static FileReader openfile()
          String line,name,file="test.txt";
          int vsize,k1,i1=0;
          boolean flag;
          Object e1,e2;
          //Vector v = new Vector();
          PriorityQueue q=new PriorityQueue();
          try
               FileReader fr=new FileReader(file);
               BufferedReader inFile = new BufferedReader(fr);
               String del=",";
               line=inFile.readLine();
               StringTokenizer tokenizer=new StringTokenizer(line,del);
               while(line!=null)
                    tokenizer = new StringTokenizer(line,del);
                    k1 = Integer.parseInt(tokenizer.nextToken());
                    e1 = tokenizer.nextToken();
                    //System.out.println(line);
                    q.InsertItem(line,i1);
                    line=inFile.readLine();
                    i1++;
               flag=q.isEmpty();
               System.out.println(flag);
               vsize=q.size();
               System.out.println(+vsize);
               e2=q.last();
               System.out.println(e2);
          }//try
          catch(FileNotFoundException exception)
               System.out.println("The File "+file+" was not found");
          catch(IOException exception)
               System.out.println(exception);
          return null;
     }//filereader
//The other file is
import java.util.*;
public class PriorityQueue
     private Object[] a;
     Vector v = new Vector();
     //Constructor
     public PriorityQueue()
     public int size() {return v.size();}
     public boolean isEmpty() { return (v.size())==0;}
     public void InsertItem(Object e,int k)
          v.insertElementAt(e,k);
          //new code
          //for(int i=0;i<size;i++)
     public Object last()
          return v.lastElement();

Why not let Java do the work - use a java.util.TreeSet instead of a Vector.
Make your PQItem implement Comparable, write a compareTo(Object o) method (and an equals method for completeness) and you're golden. Then when you add(pqItem) it goes in the exact spot it belongs.

I am trying to softproof an image using a CMYK .icc file. I sent an image from LR 5 to PS CC 2014, opened the Camera Raw FIlter, but the hyperlink to access workflow is not showing up in the CR dialogue box... Any ideas why this might be?

I am trying to softproof an image using a CMYK .icc file. I sent an image from LR 5 to PS CC 2014, opened the Camera Raw FIlter, but the hyperlink to access workflow is not showing up in the CR dialogue box... Any ideas why this might be?

I am trying to softproof an image using a CMYK .icc file. I sent an image from LR 5 to PS CC 2014, opened the Camera Raw FIlter, but the hyperlink to access workflow is not showing up in the CR dialogue box... Any ideas why this might be?

Capture an image using the web camera from a web application

Hi All,
Could anyone please share the steps what have to be followed for capture an image using the web camera from a web application.
I have a similar requirement like this,
1) Detect the Webcam on the users machine from the web application.(To be more clear when the user clicks on 'Add Photo' tool from the web application)
2) When the user confirms to save, save the Image in users machine at some temporary location with some unique file name.
3) Upload the Image to the server from the temporary location.
Please share the details like, what can be used and how it can be used etc...
Thanks,
Suman

1) Detect the Webcam on the users machine from the web application.(To be more clear when the user clicks on 'Add Photo' tool from the web application)There's not really any good way to do this with JMF. You'd have to somehow create a JMF web-start application that will install the native JMF binaries, and then kick off the capture device scanning code from the application, and then scan through the list of devices found to get the MediaLocator of the web cam.
2) When the user confirms to save, save the Image in users machine at some temporary location with some unique file name.You'd probably be displaying a "preview" window and then you'd just want to capture the image. There are a handful of ways you could capture the image, but it really depends on your situation.
3) Upload the Image to the server from the temporary location.You can find out how to do this on google.
All things told, this application is probably more suited to be a FMJ (Freedom for Media in Java) application than a JMF application. JMF relies on native code to capture from the web cams, whereas FMJ does not.
Alternately, you might want to look into Adobe Flex for this particular application.

Downloading images used to go automatically to the last folder used to save images, even when reopening Firefox. In just the past few days, possibly coinciding wih the most recent updates, the folder defaults to the 'Downloads' folder. This is most annoyi

Downloading images used to go automatically to the last folder used to save images, even when reopening Firefox. In just the past few days, possibly coinciding wih the most recent updates, the folder defaults to the 'Downloads' folder. This is most annoying. What happened? Internet Explorer 8.0 did the same thing. This was one of the reasons why I started using Firefox to download all images. I went into Tools>Options and it only lets me set another folder. I dont want to set a specific folder I want it to always go to the last folder used. So what gives?
== This happened ==
Every time Firefox opened
== possibly when the most recent updates were installed, a few days ago

Thanks jscher 2000. I guess I didn't make it clear. "It restarts with all the addons activated, and resumes with the tabs that were open before closing it." IE, it's running fine now with all the extensions activated. Everything is OK now.
So something in the Firefox code was causing the bad behavior. It's not essential that I find out what the problem was - I'm just curious. And if anybody else has this same problem, it might be nice to have it corrected at the source.

Comparing images using vector

Similar Messages

Maybe you are looking for