OpenCV python got update EXANE

[ByHnDTr]

opencv got update and its %30 more efficient on image scaning which means it could help about background pixel bot like exobot. EXANE PLEASE UPDATE!!!

[PNTX]

that was not the reason he stopped. but u could have pmd him. or something else instead of creating this thread.

[exane^]

Donate goal == 500€

[cron1003]

soon

[ByHnDTr]

Originally Posted by exane^

Donate goal == 500€

you are my favourite

Originally Posted by PNTX

that was not the reason he stopped. but u could have pmd him. or something else instead of creating this thread.

there is lots of developers here i wanted let them see it to.

[ssamko]

Thank you for interesting info...but OpenCV is still much slower than pure pixel/image search made by pixel comparison.

[Odang]

Originally Posted by ssamko

Thank you for interesting info...but OpenCV is still much slower than pure pixel/image search made by pixel comparison.

What is the difference between openCV methode to compare pixel/images and "pure" pixel/image search(what do you mean with pure?)? Shouldnt they all use the same way of comparisom(template matching)?

[ssamko]

Originally Posted by Odang

What is the difference between openCV methode to compare pixel/images and "pure" pixel/image search(what do you mean with pure?)? Shouldnt they all use the same way of comparisom(template matching)?

I have used openCV templateMatch in past (on CPU as well as GPU) and speed of finding object on full-hd pic was about 40-400ms (depends on maaany settings)...with simple ImageSearch which go 1 by 1 in pixel array (and comparing just simple pixel values) you can get match in 4-40ms.
The main difference there is in "match"..openCV has some more comlicated algorithms for that which take more time.

Here is explained OpenCV templateMatch:

PS: I havent work with OpenCV for about 2 years now...maybe some really significant improvements were made.

[ByHnDTr]

Originally Posted by ssamko

I have used openCV templateMatch in past (on CPU as well as GPU) and speed of finding object on full-hd pic was about 40-400ms (depends on maaany settings)...with simple ImageSearch which go 1 by 1 in pixel array (and comparing just simple pixel values) you can get match in 4-40ms.
The main difference there is in "match"..openCV has some more comlicated algorithms for that which take more time.

Here is explained OpenCV templateMatch:

PS: I havent work with OpenCV for about 2 years now...maybe some really significant improvements were made.

This is an AWESOME diffrence. thank "you" for interesting that info