Huge News!Announcing our $40M Series B led by Abstract Ventures.Learn More
Socket
Sign inDemoInstall
Socket

camera-zwo-asi

Package Overview
Dependencies
Maintainers
1
Alerts
File Explorer

Advanced tools

Socket logo

Install Socket

Detect and block malicious and high-risk dependencies

Install

camera-zwo-asi

python wrapper for ZWO astronomical cameras

  • 1.0.4
  • PyPI
  • Socket score

Maintainers
1

CAMERA-ZWO-ASI

python wrapper over ZWO astronomical cameras

What is it

camera-zwo-asi is a python wrapper of the C++ SDK as provided by ZWO. It provides a python object oriented interface for interacting with ZWO astronomical cameras. The version of SDK currently binded is 1.24. You may find the original C++ SDK binaries and documentation here.

Requirements

camera-zwo-asi has been tested only with :

  • python 3.10 on Ubuntu 20.04
  • python 3.9 on raspberry pi 3 (PI OS Lite 32-bit)

but is likely to work with other recent version of python3 / ubuntu / raspberry. Compilation on other linux based platforms is less likely to be successful.

Dependencies

The following APT dependencies are required:

apt install -y libusb-1.0-0-dev libgl1-mesa-glx  libglib2.0-dev libopencv-dev python3-dev cmake ninja-build libusb-dev

For raspberry, also install:

apt install -y libatlas-base-dev

Installation

from PyPI

pip install camera-zwo-asi

from source

git clone https://github.com/MPI-IS/camera_zwo_asi.git
cd camera_zwo_asi
pip install .

allow your computer to reach your USB camera

Run:

zwo-asi-udev

and follow the instruction printed on screen.

running unit-tests

Tests requires to have a usb camera plugged in. After installation from source:

cd camera_zwo_asi
pytest ./tests/test.py

Command line usage

The following command line tools are provided:

Information about connected camera(s)

# will print information about connected cameras
zwo-asi-print

Dumping the current configuration of the camera

# will create in the current folder a file called 'zwo_asi.toml' which
# contains the current configuration of the camera
zwo-asi-dump

Taking pictures

# Takes a picture and display it. If there is a file 'zwo-asi.toml' in the current
# directory, configure first the camera using it.
zwo-asi-shot

# Same as above, but does not display the image.
zwo-asi-shot -silent

# Same as above, and also save the file to /tmp/img.bmp
# For the list of supported file formats:
# https://docs.opencv.org/2.4/modules/highgui/doc/reading_and_writing_images_and_video.html#imread
zwo-asi-shot -silent --path /tmp/img.bmp

# Same as above, but ignores the 'zwo-asi.toml' file that may be in the
# current directory.
zwo-asi-shot -silent -noconfig --path /tmp/img.bmp

# Same as above, but uses the second camera (index 1).
# Will work only if at least two cameras are connected.
zwo-asi-shot -silent -noconfig --path /tmp/img.bmp --index 1

# Getting info:
zwo-asi-shot -h

Taking pictures with the configuration of your choice

# generate the file corresponding to the current camera's configuration
zwo-asi-dump

#
# edit zwo_asi.toml to your liking. 
#

# take a picture with your desired configuration
zwo-asi-shot

When setting up the configuration (by editing zwo_asi.toml), some controllable supports 'auto mode' and some are not writable. When calling zwo-asi-print, you may see which ones, by looking at the column 'auto-mode' and 'writable'. For example:

(asi sdk: 1, 24, 0, 0)
ZWO ASI294MC Pro (id: 0)
max heigth: (2822) | max width: (4144) |
colored: * | mechanical shuttger: - | st4 port: - | 
has cooler: * | is usb3 host: * | is usb3: * | 
is triggered camera: - | bayer pattern: RG
supported bins: 1 2 3 4 
supported image types: raw8 rgb24 raw16 y8 
pixel size (um): 4.63 | elec per ADU: 0.399 | bit depth: 14

|controllable               |value   |min value |max value    |auto-mode  |in auto-mode  |writable
--
|AutoExpMaxExpMS            |30000   |1         |60000        |           |              |*
|AutoExpMaxGain             |285     |0         |570          |           |              |*
|AutoExpTargetBrightness    |100     |50        |160          |           |              |*
|BandWidth                  |80      |40        |100          |*          |*             |*
|CoolPowerPerc              |0       |0         |100          |           |              | 
|CoolerOn                   |0       |0         |1            |           |              |*
|Exposure                   |10000   |32        |2000000000   |*          |              |*
|Flip                       |0       |0         |3            |           |              |*
|Gain                       |200     |0         |570          |*          |              |*
|HighSpeedMode              |0       |0         |1            |           |              |*
|MonoBin                    |0       |0         |1            |           |              |*
|Offset                     |8       |0         |80           |           |              |*
|TargetTemp                 |0       |-40       |30           |           |              |*
|Temperature                |0       |-500      |1000         |           |              | 
|WB_B                       |95      |1         |99           |*          |              |*
|WB_R                       |52      |1         |99           |*          |              |*

BandWidth supports auto-mode and is currently in auto-mode, Exposure supports auto-mode but is not currently in auto-mode. Temperature is not writable.

To set in zwo_asi.toml a controllable in auto mode, use the string "auto". For example, this sets BandWidth to auto-mode:

[controllables]
AutoExpMaxExpMS = 30000
AutoExpMaxGain = 285
AutoExpTargetBrightness = 100
BandWidth = "auto"
CoolerOn = 0
Exposure = 10000
Flip = 0
Gain = 200
HighSpeedMode = 0
MonoBin = 0
Offset = 8
TargetTemp = 0
WB_B = 95
WB_R = 52

There are also a few rules that has to be respected when setting the values of the ROI (Region Of Interest):

  • the width and height are positive, and below the max allowed values
  • the width must be a multiple of 8
  • the height must be a multiple of 2
  • binned sensor width and height must be respected (i.e. start_x + width < max_width / number bins; and start_y + height < max_height / number_bins)

For example, this respects these rules:

[roi]
start_x = 0
start_y = 0
width = 4144
height = 2822
bins = 1
type = "raw8"

note:

when the camera closes, it restores its configuration. Therefore calling zwo-asi-print after taking a picture with zwo-asi-shot may not display the configuration that was used to take the picture.

API usage

from pathlib import Path
import camera_zwo_asi

# connecting to the camera
# at index 0
camera = camera_zwo_asi.Camera(0)

# printing information in the
# terminal
print(camera)

# changing some controllables
# (supported arguments: the one that are
# indicated as 'writable' in the information
# printed above)
camera.set_control("Gain",300)
camera.set_control("Exposure","auto")

# changing the ROI (region of interest)
roi = camera.get_roi()
roi.type = camera_zwo_asi.ImageType.rgb24
roi.start_x = 20
roi.start_y = 20
roi.bins = 2
roi.width = 480
roi.height = 340
camera.set_roi(roi)

# saving this updated configuration to a file
conf_path = Path("/tmp") / "asi.toml"
camera.to_toml(conf_path)

# taking the picture
filepath = Path("/tmp") / "asi.bmp"
show = True
# filepath and show are optional, if you do not
# want to save the image or display it
image = camera.capture(filepath=filepath,show=show)

# getting a flat numpy array with the image data
image.get_data()

# getting a shaped numpy array with image data
image.get_image()

# showing the image
image.display(resize=0.25)

# saving the image (once more, for demo)
image.save(filepath)

# configuring the camera using a configuration file
camera.configure_from_toml(conf_path)

# taking a picture, overwriting the previous image
camera.capture(image=image)
image.display(resize=1.5)

Other project

python-zwoasi is another python wrapper for ZWO cameras.

Vincent Berenz, Max Planck Institute for Intelligent Systems, Empirical Inference Department

Copyright 2022 Max Planck Gesellschaft

FAQs


Did you know?

Socket

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install

Related posts

SocketSocket SOC 2 Logo

Product

  • Package Alerts
  • Integrations
  • Docs
  • Pricing
  • FAQ
  • Roadmap
  • Changelog

Packages

npm

Stay in touch

Get open source security insights delivered straight into your inbox.


  • Terms
  • Privacy
  • Security

Made with ⚡️ by Socket Inc