gesture-recognition | : hand : Recognizing `` Hand Gestures '' using OpenCV and Python | Machine Learning library

I know this sounds stupid and I'm propably very late to the party but here's the thing I want to program an gesture recogniction application (in the likes of this Hand detection or this actual finger detection) for the Xbox 360 Kinect. SDK (version 1.8) is found, installed and works, preliminary research is done - I only forgot to look in which language to write the code. The link from the SDK to the documentation would be the first thing to do but is a dead end, unfortunately.
From the provided examples it seems either to be C++ or C# although some old posts also claim Java. My question is: Is there a documentation not tied to the SDK and which pitfall are there in regard to developing in this specific case under C++/C#/Java? A post from 2011 barely covers the beginning.

Addendum: On further looking I was prompted for the Samples site from the developer toolkit - which can be reached, yet all listed and linked examples are dead ends too.

Addendum: For reference I userd this instruction - ultimately proving futile.

Found an version of NiTE here

I am working on a dataset that is a collection of 5 Hand made letters. I've uploaded the DB on Kaggle and if anyone wants to give it a look, please do.

https://www.kaggle.com/shayanriyaz/gesture-recognition

Currently, I've trained and tested several models but I keep getting 100% accuracy.

Here's my code.

I'm working with OpenCV on hand detection. But I'm struggling when trying to contours of the threshed image. findContour will always try to find white area as contour.

So basically it works in most cases but sometimes my threshed image looks like this :

_, threshed = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY|cv2.THRESH_OTSU)

So to make it works I just need to change the threshold type cv2.THRESH_BINARY_INV.

_, threshed = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY_INV|cv2.THRESH_OTSU)

And it works well.

My question is how can I determine when the threshold need to be reversed ? Does I Need to always found contours on both threshed images, and compare the result (I this case How ?) ? or there is a way to allays knows if contours are not totally missed.

EDIT : There is a way to be 100% sure contour looks like a hand ?

EDIT 2 : So I forgot to mention that I'm trying to detect fingertips and defects using this method so I need defects, which with the first threshed image I can't find them, because it reversed. See blue point on the First contour image.

Thanks.

I'm trying to get the following script to work. I found it from this website but I keep getting the error :

contours, hierarchy = cv2.findContours(thresh1,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE) ValueError: too many values to unpack (expected 2)

I have little experience in Python and OpenCV so if anyone here can help that would be much appreciated! I'm running Mac OS X 10.13.4, OpenCV 3.4.1, Python 3.6.5 and

Here is the script I'm trying to get work :