2019-05-10 01:01:39 -0600 | received badge | ● Popular Question (source) |
2018-03-08 10:46:47 -0600 | received badge | ● Good Question (source) |
2015-12-26 02:18:50 -0600 | received badge | ● Famous Question (source) |
2015-02-03 03:48:07 -0600 | received badge | ● Notable Question (source) |
2014-06-24 04:06:30 -0600 | commented question | Python: CalcEMD2 error with cost matrix I have exactly the same problem. Did you find a solution? |
2014-06-24 04:05:14 -0600 | received badge | ● Supporter (source) |
2014-03-29 08:37:01 -0600 | commented question | cv2.calcHist - sum of histogram does not equal pixel count Just confirmed berak's comment. Values for all 3 channels are correct! Thanks! It's yet not clear to me why it is implemented that way, but since it is similar to Numpy, there seems to be a reason - even if it's just to be similar to numpy... @berak: Kudos! Would you like to give the answer? |
2014-03-29 07:48:45 -0600 | commented question | cv2.calcHist - sum of histogram does not equal pixel count Just noticed: the values of the last Numpy histogram bins of the other channels are all zero. This explains why the OpenCV results for the green and blue channel are not influenced. |
2014-03-29 07:40:28 -0600 | commented question | cv2.calcHist - sum of histogram does not equal pixel count added further debugging output to the question body |
2014-03-28 09:41:26 -0600 | received badge | ● Popular Question (source) |
2014-03-28 09:40:04 -0600 | received badge | ● Editor (source) |
2014-03-28 09:37:43 -0600 | asked a question | cv2.calcHist - sum of histogram does not equal pixel count Hi, I just observed a strange behaviour of cv2.calcHist. When calculating the sum of all bin-values it sometimes does not add up to the number of pixels of an image. The number of pixels in each channel should be 262144, which is correct for the blue and green channel. The value of the red channel differes by 112. This seems to be a serious bug at least in the Python wrapper. I currently have no time to reproduce this in C++. I am using OpenCV-Python 2.4.8 (64Bit) compiled for Windows by Christoph Gohlke: http://www.lfd.uci.edu/~gohlke/pythonlibs/#opencv and Numpy 1.7.1 The error remains also after updating to Opencv 2.4.8.1 and Numpy 1.8.1 The image used was: http://upload.wikimedia.org/wikipedia/en/2/24/Lenna.png edit 2014-03-29: Some more debug data. The following two histograms have been created for the red channel. The first one is the result of the OpenCV function, the second was calculated using Numpy. OpenCV: Numpy: (more) |
2013-08-13 14:45:28 -0600 | received badge | ● Nice Question (source) |
2013-01-09 18:10:34 -0600 | received badge | ● Student (source) |
2012-08-31 09:54:08 -0600 | asked a question | cv2.VideoCapture: Cannot read from file Hi! I try to process videos using the Python interface. I downloaded OpenCV 2.4.2 and copied the cv2.py file to the Python27\Lib\site-packages folder on Windows 7. Using the following code, I'm not able to read video files. capture.isOpened() always returns false. Reading images from Webcam using cv2.VideoCapture(0) works fine, but not from files. Did I miss to copy some importtant files to the Python directory? Cheers, Sly |