-
The current feature detection algorithms struggle with very dim targets. Improve detection of dim targets.
Also look into whether it is really necessary to apply an adaptive threshold to a greyscal…
-
Summary:
I'm preprcessing the image using Brad Larsen AdaptiveThresholding. I want to stop the tesseract's default processing (Otsu thresholding). Is there any way to achieve that?
Tess-two versio…
-
Check link 3 for a specific implementation.
Relevant doc:
1. https://docs.opencv.org/master/d7/d1b/group__imgproc__misc.html#ga72b913f352e4a1b1b397736707afcde3
2. https://docs.opencv.org/java/3.0…
-
Eyegrade cannot detect exams with OpenCV 3.2, probably because of some API change with respect to OpenCV 2.4. Code in eyegrade/detection.py should be debugged to find the issue.
Commit 627afc1ed339…
-
- **median works** better than mean values
` cv2.threshold(img,np.median(img),255,cv2.THRESH_BINARY)` works better than ` cv2.threshold(img,np.mean(img),255,cv2.THRESH_BINARY)`
- mean threshold
`c…
-
Currently the code is intended for light backgrounds. If an image with a dark background is detected, the program detects a single rectangular contour the size of the image.
To fix this, we need to m…
-
reference note in Anssi Nurminen's master's thesis, "An edge in an image is defined as an above-threshold change in intensity value of neighboring pixels. Choosing a threshold value too high, some o…
-
From Wikipedia:
> The simplest thresholding methods replace each pixel in an image with a black pixel if the image intensity is less than some fixed constant T, or a white pixel if the image intens…
dzach updated
6 years ago
-
I got an error while installing opencv4nodejs on Windows 10.
I think the error was raised while building opencv4nodejs using node-gyp.
windows-build-tools were successfully installed.
And thi…
owlur updated
6 years ago
-
Hi, Developer,
I am working on RedHat 6.9 x86_64 with python/2.7.x-anaconda, ImageMagic 7.0.7-4 Q16 and GraphicsMagick 1.3.27.
I tried to pip install pgmagick and manual source install, but came…