Optical motion captue for Reach-to-Grasp movement

This project is focused on how to use motion capture in performing the RTG test that normally does with Electro-Magnetic device like 3D Guidance from NDI

OMC.py

Optical Motion Capture class

this class perform hand motion capture from video file

Inputs	Type	Description
videoPath	str	absolute or relative path to video file
videoSize	None or Tuple	if None then no video resize done in this class. if (width, height) then this class will resize output of every frame
skip	int	skip first n-th frame of video
QSize	int	how many frame you want this class to keep in memory if too many than it might flood your RAM

Methods	Description
getFrame	return next frame, if specified guarantee = True then it will wait for frame, if guarantee = False then if there is no frame in Q return None default is True
handDetect	return next frame with hand landmarks write into frame
close	properly close all of the video objects and threads must call once done using class object

---

dlt.py

Direct Linear Tranform class

you can use this class to triangulate coordinate in 2D space into coordinate in 3D space by using 2 images from different views of the object

Inputs	Type	Description	Note
xyz	array-like object	coordinates of the point on object in 3D space	There must be at least 6 calibration points for calibration process
uvL	array-like object	coordinates of the point on object left-image in 2D space	There must be at least 6 calibration points for calibration process
uvR	array-like object	coordinates of the point on object right-image in 2D space	There must be at least 6 calibration points for calibration process

Methods	Description
calibrate	calculate linear transformation of coordinates. Must call this method first before you can use getXYZ method
getXYZ	given uvL and uvR of a single point this method will return XYZ coordinate of real world coordinate

---

coor.py

Coordinate utility

this file enabled you to calibrate and use DLT easier with some helper function

Function	Description	Input	Type	Description
findBestDLT	get the best-calibrated parameters for DLT	l_file	str	path to image from left view containing chessboard
		r_file	str	path to image from right view containing chessboard
		theta	float	chessboard's tilted angle relative to floor in degrees
		phi	float	chessboard's rotated angle relative to user defined x-axis in degrees
		sqSize	float	chessboard tile size should be in mm unit
		silent	bool	operate in silentness not print anything in to terminal
		num_iterate	int	number of times to randomly select the point to find the best-calibrated parameters for DLT
EuclidDistance	get distance between two point	xyz	array-like object	coordinate of the first point on object in 3D space either from DLT or measurement
		uvw	array-like object	coordinate of the second point on object in 3D space either from DLT or measurement
save_to_file	save DLT object into file for later use	obj	DLT object	DLT object that already calibrated
		filename	str	filename to be save
load_from_file	load ready-to-use DLT object from file	filename	str	filename to be load from