You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
38 lines
1.2 KiB
38 lines
1.2 KiB
# How the inverse kinematics is calculated
|
|
|
|
|
|
|
|
The concept uses the following assumptions:
|
|
* The degrees-of-freedom are limited to
|
|
* Shoulder joint with 3dof and restrictive joint limits
|
|
* The elbow with only 1dof
|
|
* The wrist with 3dof and restrictive joint limits
|
|
* for humans, within the reachable elbow positions, some areas are more likely to be used due to less muscle strain
|
|
|
|
## Proposal 1 for am pose estimate
|
|
|
|
Input:
|
|
* controller position
|
|
* shoulder position estimate
|
|
|
|
Output:
|
|
* Elbow position
|
|
* Shoulder backslash (required to improve shoulder position estimate)
|
|
|
|
### Step 1: Initial elbow position estimate
|
|
|
|
We try to get a first good estimate using a very simple method: given
|
|
* a vector *SpineShoulder*, going from the spine at approx. breast-hight to the shoulder position, and
|
|
* a vector *ShoulderWrist*, going from the shoulder to the wrist position, which can be easily inferred from the controller position,
|
|
* calculate the cross-product like this for the **right shoulder**:
|
|
```var elbow_dir = SpineShoulder.cross(ShoulderWrist)```.
|
|
|
|
### Step 2: Consider controller orientation
|
|
|
|
|
|
### Step 3: Consider shoulder orientation
|
|
|
|
|
|
### Step 4: Determine shoulder backslash for shoulder position estimation
|
|
|