Skip to content

Latest commit

 

History

History
131 lines (101 loc) · 4.55 KB

README.md

File metadata and controls

131 lines (101 loc) · 4.55 KB
Raw

nimbus-python

Python bindings for nimbus. These bindings are ment for a remote connection (raspberry to a desktop machine).

Quick start

The following snippet connects to the raspberry and gets image data.

from nimbusPython import NimbusClient
cli = NimbusClient.NimbusClient("192.168.0.69")
header, (ampl, radial, x, y, z, conf) = cli.getImage(invalidAsNan=True)

Installation

nimbus-python uses websockets, which requires python 3.6 or higher!

pip install nimbus-python

Prerequisites

Download the current image from https://cloud.pieye.org/index.php/s/c2QSa6P4wBtSJ4K which contains nimbus-userland and all necessary linux drivers.

Getting image data

The following snippet connects to the raspberry and gets the image data.

from nimbusPython import NimbusClient
cli = NimbusClient.NimbusClient("192.168.0.69")
header, (ampl, radial, x, y, z, conf) = cli.getImage(invalidAsNan=True)

The matrices x,y,z represent a point cloud. Those can be visualized by:

The matrices have the following meaning

Matrix Explanation
ampl signal strength of each pixel
radial radial distance of each pixel to the camera center
x,y,z 3D Point cloud
conf confidence information of each pixel (valid, underexposured, saturated, asymmetric)

You can change the exposure of the Nimbus 3D. By default, auto exposure with HDR is activated. If there is a lot of movement, it may be necessary to disable HDR and use the normal auto exposure. However, it is also possible to set the exposure time manually with and without HDR.The following snippet contains the possible configurations.

# automatic exposure 
cli.setExposureMode(NimbusClient.AUTO_HDR)
cli.setExposureMode(NimbusClient.AUTO)
cli.setAmplitude(1000)  #<-- to change the desired amplitude (0 - ~5000)

# manual exposure 
cli.setExposureMode(NimbusClient.MANUAL_HDR)
cli.setExposureMode(NimbusClient.MANUAL)
cli.setExposure(5000)  #<-- to change the exposure time (0 - 65535)

If you are interested in the amount of valid, under exposured etc. pixels, you can use the following snippet as an example.

header, (ampl, radial, x, y, z, conf) = cli.getImage(invalidAsNan=True)
numUnderExposured = len(conf[conf==NimbusClient.ConfUnderExposured])
numOverExposured = len(conf[conf==NimbusClient.ConfOverExposured])
numAsymmetric = len(conf[conf==NimbusClient.ConfAsymmetric])
numValid = len(conf[conf==NimbusClient.ConfValid])

Based on this information you probably want to change the illumination time (increase the illumination in case of many underexposured pixels):

rv, data = cli.getExposure()
if rv == 0:
    # increase illumination time by 10%
    newExposure = int(data["exposure"] + data["exposure"]*0.1)
    rv = cli.setExposure(newExposure)
    assert rv==0

The illumination time can have any value between 0 and 65535.

Similarily if you want to decrease the number of frames taken by the camera, you can set a framerate value (0 means no pause at end of frame, 65535 means maximum pause at end of frame)

# fast acquisition
rv = cli.setFramerate(0)
assert rv==0
# slow acquisition
rv = cli.setFramerate(65535)
assert rv==0

Enable raw data

To enable raw data streaming, you can use the following code snippet.

from nimbusPython import NimbusClient
import time
cli = NimbusClient.NimbusClient("192.168.0.69")
cli.setExposureMode(NimbusClient.MANUAL)
cli.setExposure(5000)  #<-- to change the exposure time (0 - 65535)
cli.enaRawMode(True)
time.sleep(1) #<-- there might be still some 3D images in the pipeline
header, img = cli.getImage()
imgType = int(header[NimbusClient.HeaderImgType])
if imgType == NimbusClient.NimbusImageRaw:
    print ("raw images received")
else:
    print ("something went wrong...")
    #-> repeat image readout (cli.getImage())

Authors

Markus Proeller

See also the list of contributors who participated in this project.

License

This project is licensed under the GPLv3 License - see the LICENSE file for details

3rd party libraries

We use the following 3rd party libraries: