Hand collision with trigger object in Unity?!?

[Dnulke]

Hello! I was wondering how you call the hands in unity as they were objects? All i want to do is to make a boolean value true once the hands collide with a trigger object. If I were to do this with a simple sphere it would look like this:

void OnTriggerEnter(Collider other){
if(other.gameObject.name == "Sphere"){
isWarm = true;
}
}

Now... I don't know what to replace "Sphere" with.
The problem is that there is no hand objects in the unity hierachy until the leap motion tracks your hands in the game mode.

Thanks in advance <3

[weeba2933]

Hi Dnulke,

All hands have 3 colliders per finger, and 1 collider for palm (Total of 16 colliders).
All finger colliders have a HandModel component attached to its parent's parent (Except for Palm).

The hierarchy looks like:

HandModel
| FingerModel (Index)
|| Bone #1 (Collider)
|| Bone #2 (Collider)
|| Bone #3 (Collider)
| FingerModel (Middle)
|| Bone #1 (Collider)
|| Bone #2 (Collider)
|| Bone #3 (Collider)
| ... (Repeat for Ring, Pinky, and Thumb)
| Palm (Collider)

So what you can try is:

private bool IsHand(Collider other)
{
  if (other.transform.parent && other.transform.parent.parent && other.transform.parent.parent.GetComponent<HandModel>())
    return true;
  else
    return false;
}

void OnTriggerEnter(Collider other) 
{
  if (IsHand(other))
  {
    Debug.Log("Yay! A hand collided!");
  }  
}

Similarily, if you want to keep track of the Hand after colliding or to do something with it (Example: Change its colors), you can try:

private HandModel GetHand(Collider other)
{
  if (other.transform.parent && other.transform.parent.parent && other.transform.parent.parent.GetComponent<HandModel>())
    return other.transform.parent.parent.GetComponent<HandModel>();
  else
    return null;
}

void OnTriggerEnter(Collider other) 
{
  HandModel hand_model = GetHand(other);
  if (hand_model != null)
  {
    Debug.Log(hand_model.GetPalmPosition()); // Give me pal position of the hand! (Vector3)
  }  
}

Cheers,
Wilbur

[Dnulke]

Thanks for the reply! (And sorry for my late reply as well)

It works like a charm! So thanks a lot!

However, I have another question. This is not that related to this topic, but more about the Tracking Model in your "C# and Unity Developer guide" ( https://developer.leapmotion.com/documentation/csharp/devguide/Leap_Tracking.html ).

I can see that classes and functions related to "bone" and "finger" is used, but where are they used exactly in Unity? I have my "ControllerSandbox" object with the HandController script, but I can't seem to figure out where the classes and functions are defined. I don't know if this sounds stupid, but help would be appreciated!

Thanks in advance and thanks for the help and the wonderful product you have created.

Sincerly,
Dnulke

[weeba2933]

Hi Dnulke,

They're being used in FingerModel.cs.

The HandController is responsible for instantiating and destroying HandModels.

HandModel is responsible for creating the transformations and rotations between each fingers.

FingerModel is responsible for getting all the relative finger and bone attributes that you mentioned in Unity scale (i.e. GetTipPosition()).

You can check out these scripts in LeapMotion/Scripts/Hands.

I hope this helps!

Cheers,
Wilbur

[Dnulke]

Thank you so much for the quick asnwer!

Now it makes more sense.

One more thing It says that the Leap Motion software combines its sensor data with an internal model of the human hand. What kind of internal model is that? I mean, I don't quite understand how the Leap Motion tracks the hand in the first place. Is this by some image processing of the infrared images or how is it done?

And one very last thing! It says the Leap Motion works best when there is a high-contrast view of an object's silhouette. My group and I at our university have made a setup of the Leap Motion 'looking' downwards to a plate covered in black acrylic paint. We found that the best contrast from the hands to an object is to paint it black. Is this because of the infrared light does not reflect so well off of the the material or is it for some other reason?

Once again, thanks for your help!

Sincerly,
Dnulke

[Joe_Ward]

The internal model is a mathematical model of the hand skeleton (more or less) with constraints to help prevent the reported hand data from assuming impossible or unnatural positions. The base data is derived from stereo IR cameras.

To your second question, I'd say the answer is yes, IR light evidently doesn't reflect well off your paint. You can use the images shown in the Visualizer to see what the scene looks like from the device pov.

[Dnulke]

Thanks a lot for your answer. It makes sense. I think I got a hold of it all now!

The help is much appreciated!

Cheers!

[zanny665]

Hi,

I was reading through this code and I understand it but I am unsure of where this goes? Does it go in a script that already exists or does it go in a new script entirely?

[raminixds]

Does it still work like this @Dnulke ? Can't get it working. Do I need another Manager or Eventsystem in? Have them actually... but don't get the trigger. To check if the collider with my target object was ok, I placed a cube with a rigidbody and collider on top of it, which triggered the functions and logs...