VxW's profile - activity

Hi, yes I know the documentation. Hovever, when I have two applications (two projects) in one Solution file, have I se

setUseOptimized() Hi, I've a simple question concerning setUseOptimized() function. I'm working with Visual Studio17 a

you can try a floodfill operation

Ok, than try out the inpainting and Image Denoising function from OpenCV directly Inpainting OpenCV Image Denoising

Ok, than try out the inpainting function from OpneCV directly Inpainting OpenCV

you can try out a GraphCut RANSAC proposed by Daniel Barath and Jiri Matas at CVPR 18' Paper and GitHub. The approach is

I would try out image restauration and image inpainting like here or here but which is not a OpenCV specific problem.

Hi, you can check if the contour area is bigger than the perimeter: for contour in contours: if cv2.contourAr

Hi, you can try out some simple median filtering or oil painting Effects, which can be found here or here Best

The problem is, that Otsu do not select the optimal threshold in this case. I would suggest to try out different auto th

Hi, you can try some simple preprocessing techniques, before thresholding, but these depends mostly on the images you h

you can also try some implementations found by google. see here or here. Give attention to how often you apply the filte

you can use an anisotropic diffusion filter developed by Perona-Malik but it could be slow as well. try it out! see her

Hi, interesting problem and I think using morphological operations would not work here. I don't know if it could work

Hi, to find the width of a contour for each position is not so an easy task but there exist approaches. You can do it

Hi, to find the width of a contour for each position is not so an easy task but there exist approaches. You can do it

Hi, to find the width of a contour for each position is not so an easy task but there exists approaches. You can do it

this is also a good tutorial how to translate and rotate images correctly

Hi, an easy approach could be: 1) remove the inhomogeneous background illumination apply a big gaussian bluring and s

Hi, First I would calculate the normal vector for each white line. Afterwards, iterate for each white point along the c

Hi, First I would calculate the normal vector for each white line. Afterwards, iterate for each white point along the c

Hi, an easy approach could be: 1) remove the inhomogeneous background illumination apply a big gaussian bluring and s

Hi, an easy approach could be: 1) remove the inhomogeneous background illumination apply a big gaussian bluring and su

Hi, an easy approach could be: 1) remove the inhomogeneous background illumination apply a big gaussian bluring and su

Hi, you can try to calculate the thinnes ratio: T = 4*Pi ( Area/Perimeter^2) this measure has a maximum of 1, which cor

Hi, this link should be helpful

Hi, if you only need the area of the spherical objects I would first apply a median filter: Afterwards you can use a

Deep learning is not the solution for all. At least you will need thousand and much more images for learning which is no

Hi, you can calculate the extreme points of the contour. leftmost = tuple(cnt[cnt[:,:,0].argmin()][0]) rightmost = tu

Hi, I've found an OpenCV wrapper (cvsba) for the Sparse Bundle Adjustment library (sba), which was developed by M. Lour

Hi, try out different filters for preprocessing. result of median filter or result of variance filter: based on t

I havn't tried it but there are lot of plotting frameworks in c++ looks simple: https://github.com/lava/matplotlib-cpp

I havn't tried it but there are lot of plootting frameworks in c++ looks simple: https://github.com/lava/matplotlib-cpp

are you working in c++ or python or...?

it's the reduce function

Hi, this paper is also interesting, where they use a Harr Wavlet transform to determine bluriness and sharpness: Blur

open the exe file with the depency walker on your computer and have a look which dlls are refered http://www.dependency

I've downloaded the file and was not able to run it because opencv_video249.dll is missing. It could be that on your com

Hi, I think you have mixed up a 32bit environment with a 64bit one, so check your environment

Within a selfie the face should be always in foreground therefore, apply a face detection and select the face region as

Hi you could use center of gravity of the contour: cx = [] cy = [] pos = [] for i in range(0, len(conto

"The formula for intersection will be SI= Max(0, Min(XA2, XB2) - Max(XA1, XB1)) * Max(0, Min(YA2, YB2) - Max(YA1, YB1))

Hi, the illumination is not so important, but they should be constant (avoid interfering light). In a first step I wo

Hi, first Question: a standard calibration is necessary (like chessboard or circles grid) because you have to eliminate

Hi, first Question: a standard calibration is necessary (like chessboard or circles grid) because you have to eliminate

you also have to learn C++, see e.g. C++ Templates <uchar> is a template parameter and means that img_pyr2 must b

Hi, first Question: a standard calibration is necessary (like chessboard or circles grid) because you have to eliminate