Compiler/OPT9221: Writing calibration parameters in register OPT9221

Hello!

I am sorry for so long pause.

Config-file NoCalibration.

Thank you for your answer!

But I have the same troubles in VoxelViewer, when I try to calibrate.

I did as you advised me, choose only Non-linearity Calibration in Calibration Wizard with NoCalibration profile.

I tried both to load csv-file and to write the values by hand.

I use win7-64bit and win10-64bit.

VoxelViewer 0.9.10.

all log info:

Voxel Viewer v0.9.10

Texas Instruments Inc.

INFO: ParameterDMLParser: Found register map with name 'OPT9221' (vPG1P0)

INFO: HardwareSerializer: Received 42851 bytes from hardware in 1.9344 s

INFO: Data2DCodec: Original number of 8-bit offsets = 196

INFO: Data2DCodec: Number of invalid pixels in original data = 0

INFO: Data2DCodec: Current number of 8-bit offsets = 196

INFO: CameraSystem: Successfully refreshed parameters for TintinCDKCamera(0::0451:9105::916103915826).

INFO: ToFCameraBase: Starting with 320x240@30fpsUsing rate factor = 0.1

Using rate factor = 0.0825973295022

Using rate factor = 0.0825973295022

Using rate factor = 0.0825973295022

C:\PROGRA~1\TEXASI~1\VOXELV~1.10\VOXELV~1\pyqtgraph\graphicsItems\ImageItem.py:354: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future

INFO: DepthCamera: Streaming stopped.Lens Calibration False

Modulation Frequency Calibration True

Cross-talk Calibration True

Non-linearity Calibration True

Temperature Calibration True

Common Phase Calibration True

Pixelwise Phase Calibration True

INFO: DepthCamera: Setting parameter 'op_clk_freq'

INFO: DepthCamera: Setting parameter 'sub_frame_cnt_max'

INFO: DepthCamera: Setting parameter 'quad_cnt_max'

INFO: DepthCamera: Setting parameter 'mod_freq1'

INFO: DepthCamera: Setting parameter 'mod_freq2'

INFO: DepthCamera: Setting parameter 'intg_duty_cycle'

INFO: DepthCamera: Setting parameter 'dealias_en'

INFO: DepthCamera: Setting parameter 'unambiguous_range'

INFO: DepthCamera: Setting parameter 'mix_volt'

INFO: DepthCamera: Setting parameter 'intg_time'

INFO: DepthCamera: Setting parameter 'illum_power_percentage'

INFO: DepthCamera: Setting parameter 'delay_fb_corr_mode'

INFO: DepthCamera: Setting parameter 'delay_fb_dc_corr_mode'[2065 2065 2065 2065 2172 2479 2792 3102 3403 3731 3986 3986 3986 3986 3986

3986]

Traceback (most recent call last):

File "D:\voxelviewerbitbucket\voxelviewer-devel\build\VoxelViewer\out00-PYZ.pyz\PyVoxelViewer.calibration.CalibrationNonLinearityPage", line 289, in calibrate

IndexError: index 2 is out of bounds for axis 1 with size 2

you were clear.

in folder ~/.Voxel/conf of win7 computer  "TintinCDKCamera" and "916103915826.bin", and only "916103915826.bin" win10 computer.

file "TintinCDKCameraNoCalibration.conf" is found in folder Program files/Texas Instruments/VoxelViewer 0.9.10/VoxelViewer/voxelsdk/share/voxel-0.6.11/conf

It seem it intall by default VoxelViewer installer.

Is it means I need to move file "TintinCDKCameraNoCalibration.conf" in ~/.Voxel/conf folder?

file "TintinCDKCameraNoCalibration.conf" in application. (but renamed from .conf to .txt only here)

[global]
id = 132
name = No Calibration

[params]

[defining_params]
op_clk_freq = 0
sub_frame_cnt_max = 4
quad_cnt_max = 4
mod_freq1 = 48
mod_freq2 = 48
frame_rate = 30.00
intg_duty_cycle = 9
dealias_en = 0

[calib]
fx = 220.0
fy = 220.0
cx = 160
cy = 120
k1 = 0
k2 = 0
k3 = 0
p1 = 0
p2 = 0

1. This line in function getCoefficients:

112 offsetPoints = np.arange(0., 4096./2**(2-period)+1, 256./2**(2-period)) 

give array with 17 elements from 0 up 4096 with equidistantly step 256 when period = 2;
[ 0. 256. 512. 768. 1024. 1280. 1536. 1792. 2048. 2304. 2560. 2816. 3072. 3328. 3584. 3840. 4096.]

113 y = np.around(np.interp(offsetPoints, measuredPhase, expectedPhase))

function "interp" return array with the same size as "offsetPoints" i.e. 17 elements
docs.scipy.org/.../numpy.interp.html

and there is not any erasing processes with array elements in other places of code.

Suramya,
Thank you for your answer!

Is the method of checking calculated NonLinearity Calibration coefficients?

Do I need to make tg_dis stop/start and/or PLL_Update when writing nonlinear calibration parameters (not only this ones but other too) to the OPT9221?

Hi, I am also trying to understand how the python calibration works for OPT8241. I understand that 4 phase values are collected to calculate the distance value. But in this discussion I see only one phase being mentioned. How are 4 raw phase values connected to this single phase used in calibration conf file ?

Suramya, Thank you for your answer!

I am sorry for unclear question.

"Is the method of checking calculated NonLinearity Calibration coefficients?"

I meant if I got some array nonLinearityCalibration[16] after calibration process, than I wrote this one to register opt9221, how  can I understand that phase/depth map is better than before?

In NonLinearityCalibration.py

151    if y1:

154    print "phase_lin_coeff1 = " + data

may be some bug in line 151:

if y1.all

and some bug in line 154:

print "phase_lin_coeff1 = " + data1

and about common offset correction:

Is some main physical/technical difference between CommonPhaseOffsetSimpleAverage and CommonPhaseOffsetComplexAverage algorithms?

And in CommonPhaseOffsetSimpleAverage.py

82     centerShape = [cx-window/2, cx+window/2, cy-window/2, cy+window/2]

103 averagePhase = np.sum(averagePhase)/(window*window)

when window = 4. it is meant centsrShape is square with side 5 at center (cx, cy) (25 point)

but averaging is made by window x window, (16 points)

Or may be I do not understand something?