Skip to content
This repository has been archived by the owner on Feb 17, 2023. It is now read-only.

Latest commit

 

History

History
191 lines (142 loc) · 5.94 KB

README.rst

File metadata and controls

191 lines (142 loc) · 5.94 KB
Raw

Bullseye Laser Beam Profiler

Bullseye is a laser beam analysis application. Images can be acquired from any USB or Firewire camera using the DC1394 standard supported by pydc1394 or pyflycapture2 from pointgrey. The beam analysis mostly adheres to ISO-11146 and determines centroid, 4-sigma width (~1/e^2 intensity width), rotation and ellipticity. The user interface is build on enthought/{traits, chaco}.

Cameras

All monochrome cameras that are supported by libdc1394 (http://damien.douxchamps.net/ieee1394/libdc1394/) or libflycapture2 (http://www.ptgrey.com/support/downloads/download.asp) are suitable for laser beam profiling. That includes a lot of good and affordable firewire and USB cameras from several different manufacturers (http://damien.douxchamps.net/ieee1394/cameras/).

There are a few limitations:

  • Glass windows: they need to be removed, including the window that is typically glued to the chip. Either break it with the chip pointing downwards or pull it out while heating the epoxy. This is dangerous and will void the camera's warranty. It will also make the chip very susceptible to dust and it will age much faster. But it it necessary: otherwise fringes will negatively impact image quality. Remember that even 1% stray light due to reflections from glass-air interfaces leads to 40% peak to peak variations in intensity.
  • ND filters -- reflective or volume absorptive need to be angled significantly to keep reflections from interfering at the chip. Those filters also need to be of good quality to not distort the beam.
  • With front-illuminated silicon chips, wavelengths longer than ~1050nm penetrate deeper into the chip and lead to long living excitations and smearing of the image along the vertical shift direction.
  • Above 1100nm and below 400nm, the quantum efficiency of Silicon chips is typically below 5%. Increased powers can destroy the chip or bleach it.
  • Monochrome chips are recommended. Otherwise the Bayer color filter pattern will skew results.

Installation

Ubuntu/Debian

Get the package from the PPA:

$ sudo add-apt-repository ppa:jordens/bullseye-trunk
$ sudo apt-get update
$ sudo apt-get install python-bullseye

PyPi

Use either easy_install or pip to get and install the required packages from the python package index.

Source

Chaco, Traits and TraitsUI from enthought (http://code.enthought.com/) are available prepackaged under most GNU/Linux systems, but also come included in the highly recommended Python distributions Python(x,y) (http://www.pythonxy.com) and the Enthough Python Distribution EPD (http://www.enthought.com/products/epd.php) for Windows and Mac.

To access cameras via libdc1394, the python wrappers pydc1394 (https://launchpad.net/pydc1394) are needed. The bazaar branch is at lp:~jordens/pydc1394/work (bzr branch lp:~jordens/pydc1394/work), the Ubuntu/Debian package python-dc1394 is included in the PPA. From source, use sudo python setup.py develop to link the source tree into your python installation.

Many cameras can also be accessed using the PointGrey flycapture2 library via the Cython based wrapper pyflycapture2 (https://launchpad.net/pyflycapture2). Download the library and headers from PointGrey (http://www.ptgrey.com/support/downloads/download.asp account required, closed source) and install them. Get the wrapper source from the bazaar branch (bzr branch lp:pyflycapture2) and install it via the usual way or use sudo python setup.py develop. This needs cython.

Usage

To quickly check out Bullseye without installing, run it from the source directory with one of:

$ python -m bullseye.app
$ python -m bullseye.app -camera fc2://index/0
$ python -m bullseye.app -camera none:
$ python -m bullseye.app --help

After installing it via sudo python setup.py install or linking it with sudo python setup.py develop just run one of:

$ bullseye
$ bullseye -camera fc2://index/0
$ bullseye -camera none:
$ bullseye --help

The none: scheme uses a dummy noisy gaussian source that also works without pydc1394 or pyflycapture2 installed.

User Interface

Beam size definition http://www.rp-photonics.com/spotlight_2007_07_11.html ISO-11146 (moments) versus least squares fit versus clipping http://www.rp-photonics.com/beam_radius.html offset control via background percentile, energy inclusion radius, roi, average, and dark frame subtraction

camera settings shutter auto shutter gain framerate

processing settings dark average background ignore track

image colormap http://www.research.ibm.com/people/l/lloydt/color/color.HTM http://dx.doi.org/10.1109/MCG.2007.323435 http://www.jwave.vt.edu/~rkriz/Projects/create_color_table/color_07.pdf for an accessible copy http://www.slideshare.net/ptomato/20101007-lunch-meeting

green 2sigma, 6sigma radius markers horizonatal and vertical sum plot and gauss via moments green markers for 2sigma major and minor sum and gauss via moments green markers for 2sigma text field zooming and panning with mouse dragging and mousewheel ctrl-s to save pdf of main screen as bullseye.pdf save images: --save_format replay with -camera replay://glob/path*.npz csv output with --debug info and --log log filename

Wishlist/TODO

  • info in window title
  • leastsq for those who think they need it
  • black+jet colormap for those who think they need it
  • peak intensity
  • waist from peak curvature and peak height h*(1-x**2/sigma**2/2) in pm sigma/2
  • speed: copy-less CMapImagePlot in chaco github, speed up ranges
  • webcam/opencv (from Beams)
  • auto tracking should move view
  • float averaging or summing (from Beams)
  • group ui with tabs or menus (from Beams)
  • single-shot (from Beams)
  • re-process once when not active
  • save_format ui, save once
  • fix a/b index range to not jump
  • fix a/b rounding error