Timeline for Improving accuracy of findContour in OpenCV Python
Current License: CC BY-SA 3.0
14 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 5, 2022 at 14:27 | history | edited | Jeru Luke |
edited tags
|
|
| Dec 29, 2015 at 9:04 | answer | added | Moritz | timeline score: 3 | |
| Dec 29, 2015 at 0:49 | history | edited | batlike | CC BY-SA 3.0 |
added 209 characters in body
|
| Dec 29, 2015 at 0:47 | comment | added | batlike | The detectable contours seem to cover all the features quite well, not unlike "fill" in ImageJ. | |
| Dec 28, 2015 at 23:34 | comment | added | sturkmen |
try cv2.drawContours( original, contours,-1, (0,0,255), 2) to test drawing all contours
|
|
| Dec 28, 2015 at 23:13 | comment | added | Moritz | Personally, I do find it sometimes hard to understand the documentation of opencv. Scikit-image might be easier to work with scikit-image.org/docs/dev/auto_examples/edges/… | |
| Dec 28, 2015 at 23:13 | comment | added | batlike | Definitely a problem with detection, since it is upstream of draw contours. | |
| Dec 28, 2015 at 23:09 | comment | added | Moritz |
Is this rather a problem with the detection, or with the drawContours method ?
|
|
| Dec 28, 2015 at 23:07 | comment | added | Moritz | I do not know this method of opencv, but usually you can play around with some threshold. Probably canny edge detection. Another strategy would be to use blob detection. | |
| Dec 28, 2015 at 22:33 | history | edited | batlike | CC BY-SA 3.0 |
Added code (per request)
|
| Dec 28, 2015 at 22:26 | comment | added | batlike | Please see new edit. | |
| Dec 28, 2015 at 22:26 | history | edited | batlike | CC BY-SA 3.0 |
Added code (per request)
|
| Dec 28, 2015 at 16:52 | comment | added | sturkmen | please post your code | |
| Dec 28, 2015 at 6:00 | history | asked | batlike | CC BY-SA 3.0 |