name | 16 Jun 01:37 2012
Picon

Color cast removal

Hi,

I'm wondering whether you guys have any idea how the color correction methods of e.g.
- Photoshop: "Auto color"
- Paint.NET: "Auto-level"
(both yield similar results)

work? I'd like to implement it using OpenCV.

Note: it is NOT a simple histogram equalization or normalization applied to the individual R, G, B
channels, but it seems to be somehow a technique that aligns the 3 histograms, i.e. the peaks, but I have no
clear conception of how they do that exactly). Maybe you could point me to existing point, or some
literature that describes the procedure? Googling for color cast removal, color tint removal etc.
yields either bad results, or links to complex papers that are "over the top" w.r.t. effort for coding on my side.

Best regards!

PS: nope, GIMP does not have a filter that produces comparable results, otherwise I would've scraped their
code :D

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
mark_verlab | 17 Jun 00:31 2012

Physical lens model from camera calibration

We calibrated a camera and can get corrected images. We want to design an add-on lens for this camera. We
could do a better design job if we could include a model the lens of the current camera.

The image processing expert on the project can get the camera matrix from calibration.

The optics designer uses Zemax and does not know how to transfer these parameters into a physical model of
the camera's lens.

Does anyone know how to take the intrinsic parameters of the camera model and get the specification of an
equivalent spherical lens ? (this would include the radius of the lens curvature)

If not, has anyone heard of someone who did this ?

Cheers,
   Mark

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
Juliano Cruz | 15 Jun 19:38 2012
Picon

errors when using traincascade and haartraining



Hi folks,

I'm trying to train a Haar Cascade Classifier, but I'm experiencing some errors for either opencv_haartraining and opencv_traincascade. I have no idea what can be the cause.
I'm following exactly what some tutorials are telling me to do. (http://achuwilson.wordpress.com/2011/07/01/create-your-own-haar-classifier-for-detecting-objects-in-opencv/http://note.sonots.com/SciSoftware/haartraining.html  and   http://docs.opencv.org/doc/user_guide/ug_traincascade.html)
If some someone knows what I'm doing wrong, please let me know.

Thanks in advance...
Juliano Cruz


Here is the outputs I get:
 
[root <at> K Haar]# opencv_haartraining -data haart -bg negativoCampFut.txt -vec positivoCampFut2.vec -nstages 20  -mode ALL -w 72 -h 96 -nonsym -mem 2048
Data dir name: haart
Vec file name: positivoCampFut2.vec
BG  file name: negativoCampFut.txt, is a vecfile: no
Num pos: 2000
Num neg: 2000
Num stages: 20
Num splits: 1 (stump as weak classifier)
Mem: 2048 MB
Symmetric: FALSE
Min hit rate: 0.995000
Max false alarm rate: 0.500000
Weight trimming: 0.950000
Equal weights: FALSE
Mode: ALL
Width: 72
Height: 96
Applied boosting algorithm: GAB
Error (valid only for Discrete and Real AdaBoost): misclass
Max number of splits in tree cascade: 0
Min number of positive samples per cluster: 500
Required leaf false alarm rate: 9.53674e-07

Tree Classifier
Stage
+---+
|  0|
+---+

OpenCV Error: Insufficient memory (Failed to allocate 5618020832 bytes) in OutOfMemoryError, file /mnt/CMM/OpenCV/OpenCV-2.4.0/modules/core/src/alloc.cpp, line 52
terminate called after throwing an instance of 'cv::Exception'
  what():  /mnt/CMM/OpenCV/OpenCV-2.4.0/modules/core/src/alloc.cpp:52: error: (-4) Failed to allocate 5618020832 bytes in function OutOfMemoryError
Abortado (imagem do núcleo gravada)




[root <at> K Haar]# opencv_traincascade -data haart -bg negativoCampFut.txt -vec positivoCampFut2.vec -numStages 20  -mode ALL -w 72 -h 96 -nonsym -mem 2048

terminate called after throwing an instance of 'std::logic_error'
  what():  basic_string::_S_construct null not valid
Abortado (imagem do núcleo gravada)



__._,_.___

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com




Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required)
Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured
Visit Your Group | Yahoo! Groups Terms of Use | Unsubscribe

__,_._,___
Somnath Mukherjee | 15 Jun 19:27 2012
Picon

Looking For a Long Term or Short Term Research Project in Computer Vision & Image Processing

HI
We are a team of Research and Development actively looking for a
Research project based on Computer Vision and Pattern recognition on
the following topics:-
Brain MRI Image Segmentation, Ultrasound Image Filter, Object
Detection in RADAR Image Processing etc. we have an good experienced
regarding this topics and also other topics.
Interested Group or Team may contact with me .

--

-- 
   Somnath Mukherjee
somnath.7.mukherjee <at> gmail.com

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
Rohan | 15 Jun 19:57 2012
Picon

Light GUI based OS

Hello,
I am just getting started with OPENCV & would like to do a project on Face detection.
I want to install a light GUI based Ubuntu on my board.
Could you please suggest a very light weight Ubuntu OS that would get me started off?

Thanks.

- Rohan

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
mjlaielli | 16 Jun 02:24 2012
Picon

Why does haartraining break with too few samples?

I'm debugging a modified version of haartraining.cpp and, in certain instances, I would like to do so with
the simplest input parameters possible.  However, when I try anything below about 50 pos/50 neg samples,
an infinite for-loop occurs when the program evaluates the current cascade - specifically, when it tries
to calculate the false alarm rate.  Any ideas on why this happens?

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
ejohns85 | 16 Jun 19:37 2012
Picon

vector<Mat> problem....

Hello,

I am creating a vector of Mats. However, I am finding that when I change the value of an element of one vector,
it also changes the value of that element in all the other Mats in the vector.

For example, in the code below, a and b are both equal to "1.234", even though I have only explicitely
assigned this element in one of the matrices. It seems as if when creating a vector of Mats, they are all
shallow copies of each other, whereas I want each to be an independent Mat. How can I do this?

Thanks.

Code:

vector<Mat> images(2, Mat::zeros(10, 10, CV_32F));
images[0].at<float>(5, 5) = 1.234;
float a = images[0].at<float>(5, 5);
float b = images[1].at<float>(5, 5);

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
rgap123 | 17 Jun 01:27 2012
Picon

Algorithm to remove bright white spots

Hi,

Is there a way to remove  bright white spots(brightness) in an rgb image?

thanks.

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
rahulgaurav.kist | 17 Jun 01:52 2012

new to OpenCV; Coding guidelines needed for contour based problems

	
 	In an image which shows a binary silhouette of some biscuits which were produced in a factory. Write a
program which shows/marks the broken defective biscuits which must not be packaged.  E.g. put a circle or
rectangle to mark the poor biscuits. 

PS: in the image the proper biscuits are in kind of circle (not a perfect one though) and the improper ones are
circles with broken edges.

I am trying to use this code thinking that it will help me find the circle at least ..But it is not :(

It compiles but does not give any fruitful output just the image.

#include "stdafx.h"
#include "ml.h"
#include "cxcore.h"
#include "cxtypes.h"
using namespace std; 
using namespace cv; 
#include <cv.h>
#include <highgui.h>
#include <iostream>
#include <stdio.h>

using namespace cv;
/**  <at> function main */
int main(int argc, char** argv)
{
  Mat src, src_gray;
	IplImage* src;
	src = cvLoadImage( "biscuits.bmp");
	/*cvNamedWindow( "Example1", CV_WINDOW_AUTOSIZE );
	cvShowImage( "Example1", Img); 
	cvWaitKey(0); 
	cvReleaseImage( &Img ); */
  /// Read the image
//  src = imread("trip.jpg", 1 );

  if( !src.data )
    { return -1; }

  /// Convert it to gray
  cvtColor( src, src_gray, CV_BGR2GRAY );

  /// Reduce the noise so we avoid false circle detection
  GaussianBlur( src_gray, src_gray, Size(9, 9), 2, 2 );

  vector<Vec3f> circles;

  /// Apply the Hough Transform to find the circles
  HoughCircles( src_gray, circles, CV_HOUGH_GRADIENT, 1, src_gray.rows/8, 200, 100, 0, 0 );

  /// Draw the circles detected
  for( size_t i = 0; i < circles.size(); i++ )
  {
      Point center(cvRound(circles[i][0]), cvRound(circles[i][1]));
      int radius = cvRound(circles[i][2]);
      // circle center
      circle( src, center, 3, Scalar(0,255,0), -1, 8, 0 );
      // circle outline
      circle( src, center, radius, Scalar(0,0,255), 3, 8, 0 );
   }

  /// Show your results
  namedWindow( "Hough Circle Transform Demo", CV_WINDOW_AUTOSIZE );
  //imshow( "Hough Circle Transform Demo", src );
cvShowImage("Hough Circle Transform Demo", src_gray); 
  waitKey(0);
  return 0;
}
//show your results
  namedWindow( "Hough Circle Transform Demo", CV_WINDOW_AUTOSIZE );
  //imshow( "Hough Circle Transform Demo", src );
cvShowImage("Hough Circle Transform Demo", src_gray); 
  waitKey(0);
  return 0;
}

Waiting for some response :)

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
pattern | 16 Jun 14:05 2012
Picon

visualize ROI in a transparent color !!!

hi,
I have an ROI(not a rectangular region) circumscribed by a contour: vector<vector<Point>> cont. how can I
visualize it in a transparent color?

Thanks in advance

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com
vaclav_sal | 15 Jun 18:34 2012
Picon

How to convert MFC CPoint to OpenCV cvPoint?

Here is the simple problem for all you MFC with OpenCV gurus.
I use MFC  primarily as UI and I need to convert windows MFC CPoint to cvPoint for processing in OpenCV.
I have found some questions about it on the web , but no real answers.
Would cvTransform work?
Or could I overlay OpenCV image with my bitmap and get the cvPoint from the mouse that way?
Don't really want to do that since it took me some time to convert image to bitmap!
And I would probably mess up views updating via MFC. 
Anybody has a sample code willing to share with the group?
Thanks for all your help.
Vaclav 

------------------------------------

Change settings: http://www.yahoogroups.com/mygroups, select
   Get Emails (get all posts)
   Daily Digest (one summary email per day)
   Read on the web (read posts on the web only)Or Unsubscribe by mailing OpenCV-unsubscribe <at> yahoogroups.com

Gmane