We just received a new batch of A4 calibration pattern boards for the optical calibration of cameras. Printed on white backed aluminium laminate material for highest precision through extra stiffness and flatness. Calibration patterns are required for the optical calibration of cameras i.e. for computer vision applications. Rather simple for the visual spectrum - much more challenging in the infrared spectrum with cameras intended for thermography. We designed special calibration patterns for the calibration of our multispectral camera (infrared + visual).
Having some code at hand to generate the patterns, we decided to testdrive the dashboarding solution streamlit. We programmed a small calibrator board generator based on opencv, that you can try and use for free in the streamlit cloud: calibration pattern generator. The dashboard runs on limited resources, so please don’t over stress it. Also we found and reported 3 bugs during our evaluation - as of version 2022.2 the interactive board may still become stuck and sometime the controls don’t play well. In a production app we could probably work around the issues, but for this demo we decided to wait for the fixes to be introduced in the next streamlit version.
As for streamlit: it’s actually straight forward to use. And we would recommend it to individuals to present and make their results available to a small group of users. If you’re fluid in the python programming language and you need to present any kind of data interactively (CFD optimization results, modelling, etc) to your peers, you definitely should give it a try. Their streamlit example gallery has plenty of impressive examples.
UPDATE There are quite a few people interested in the tools. The cloud service that runs our python code may be slow to respond to your requests. You can also use these tools locally on your own computer, the source code is available on github.
As engineers we often need rough estimates to get an order of magnitude (hopefully much better!) of some physical behaviour. In fluid dynamics and heat transfer this often results in using experimental correlations. Not a problem if you have just worked on it and can remember the 8-digit constant and the general form of the equation. But two weeks later? A year later? And sometimes you’re not actually an expert on the question at hand but still need a quick estimate: These calculations can be surprisingly time consuming. We started our nuflo engineering toolbox as a collection of simple tools for this kind of task. At the same time it serves to conserve our research and know-how.
These small online tools come with a short theory section to give an overview of the influencing parameters and an interactive section. Usually, the interactive section has sliders (and fields) to enter values and a plot showing the effect of the parameter change. The online version runs on a free heroku cloud instance and may be slow to start. If you find yourself using these tools a lot, contact us, or if you feel confident you can get the source code for free on our github account.
The first tool that we put online allows to interactively estimate the momentum boundary layer thickness based on simple correlations for the laminar and the turbulent regime.
Hint: Open the tools in a separate narrow browser window and place it next to other software’s windows.