Ask Your Question

Jupiter's profile - activity

2020-01-19 03:07:41 -0600 received badge  Notable Question (source)
2019-10-15 19:36:01 -0600 received badge  Student (source)
2019-10-15 16:11:38 -0600 received badge  Popular Question (source)
2018-12-30 17:16:09 -0600 received badge  Popular Question (source)
2014-07-25 07:06:54 -0600 commented question cvHaarDetectObjects with ROI and Multithreading

I work with OpetCvSharp and it only has a shell method Cv.HaarDetectObjects, not Cv.ClassifierCascade

2014-07-25 05:30:49 -0600 asked a question cvHaarDetectObjects with ROI and Multithreading

I am using cvHaarDetectObjects against an image with ROI set in parallel threads. One thread detects eye pair, another – nose, etc. The result is wrong as the ROI is a shared resource: each thread wants to have it’s own ROI which is impossible. My question is: what can I do to enable parallel Haar object detection?

2014-07-15 06:59:24 -0600 commented question haarcascade_mcs_lefteye.xml - for which eye?
  1. Could you please give more details on how you combined the results? Did you just average them?
  2. In case you do not segregate left and right eyes, why not rather use the common detector haarcascade_eye.xml?
2014-07-15 06:48:59 -0600 commented answer haarcascade_mcs_lefteye.xml - for which eye?

Thank you!

2014-07-15 06:48:27 -0600 received badge  Scholar (source)
2014-07-15 06:48:21 -0600 received badge  Supporter (source)
2014-07-15 04:33:37 -0600 asked a question haarcascade_mcs_lefteye.xml - for which eye?

I am going to use separate haar cascades to detect left and right eyes. My question is: is haarcascade_mcs_lefteye.xml applicable to the left eye of the person on the photo or to the eye at the left from the observer’s point of view? I cannot decide on my own because both haarcascade_mcs_lefteye.xml and haarcascade_mcs_righteye.xml detect both eyes on the photos that I have tried.

2014-02-27 01:56:37 -0600 commented question I need an equalization function with mask like this: equalizeHist(src, dst, mask)

I found implementation of equalizeHist(src, dst, mask) on the net: https://github.com/hengli/camodocal/blob/cd2044ace16c0847c132582034cefacd76122f2b/src/gpl/OpenCVUtils.h

Resolved ))

2014-02-26 03:44:38 -0600 commented question I need an equalization function with mask like this: equalizeHist(src, dst, mask)

And how about using MaskROI with Mat? Can I form a non-rectangular Region Of Interest?

2014-02-26 02:36:38 -0600 asked a question I need an equalization function with mask like this: equalizeHist(src, dst, mask)

In order to be able to process complicated regions, I need a equalizeHist function with 3 arguments. But unfortunally, in OpenCV it is not implemented. Is there any workaround? Thank you.

2014-02-25 05:37:53 -0600 received badge  Critic (source)
2014-02-20 04:02:43 -0600 asked a question Excluding some areas from image processing

Is it possible to apply equalizeHist method to a doubly-connected domain? I want to equalize not the whole image but to exclude certain areas from processing leaving them intact. To do that, I am thinking of using transparent “color”. I’d like to fill the excluded areas with transparent “color” before passing the image to OpenCV. Any ideas?

2014-02-18 07:47:42 -0600 received badge  Editor (source)
2014-02-18 02:02:25 -0600 asked a question Package received by NuGet does not contain dll-files

I have installed OpenCV 2.4.8 package for a C++ VS project using NuGet package manager. The problem is that the package does not contain any dll-libraries nor any lib files. Folders x64 and x86 are missing from OpenCV.2.4.8\Build in the package as compared to a stand-along installation of OpenCV 2.4.8.

The strange thing is that different versions of package OpenCV for C++ are shipped with different folder structure. Some include dlls and libs, other do not.

OpenCV 2.3.1 is shipped as two packages: OpenCV_binaries.2.3.1 and OpenCV_headers.2.3.1, the former including all libraries, the latter including header files.

OpenCV 2.4.6-nuget has folders bin and include with libraries and header files respectively.

OpenCV 2.4.6 is shipped as two packages: OpenCV.2.4.6 and OpenCV.redist.2.4.6, the former containing lib folder, the latter containing bin folder. Lib and dll files are present.

OpenCV 2.4.7 and OpenCV 2.4.8 are shipped without bin or lib folder. So lib and dll files are missing!!!