Help researchers develop a smartphone app to measure formaldehyde in your home

[Nick Shapiro]

Hi there,  

 

We are seeking user feedback on a smartphone app. Researchers at Ohio State University, in partnership with Public Lab, are developing a smartphone app to quickly and quantitatively read low-cost formaldehyde exposure badges. We are hoping to make this app as useful and accessible as possible, and would love your feedback! Please follow this link (https://u.osu.edu/smartform/beta-test/) to download the beta version of app and follow steps to the ~10 minute user survey. Participation is free and voluntary, and you can skip questions or close the survey at any time. Thank you so much for your time!  

 

Links to beta app download:

    Android: Link

    iPhone: Link

Badge images before and after exposure

 

Best Wishes,  

The SmART-Form team

---

Nick Shapiro

Open Air Fellow

Public Laboratory for Open Technology and Science

publiclab.org  / @publiclab  / @zBoratory

[Patrick Hixenbaugh]

Do you need a printer for this?

--
Post to this group at plots-waterquality@googlegroups.com
 
Public Lab mailing lists (http://publiclab.org/lists) are great for discussion, but to get attribution, open source your work, and make it easy for others to find and cite your contributions, please publish your work at http://publiclab.org
---
You received this message because you are subscribed to the Google Groups "plots-waterquality" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plots-waterquality+unsub...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Gretchen Gehrke]

Hi Patrick! 

You can download it onto a laptop or desktop, and then take a picture of your computer screen. Or, you can print it. 

Thanks! 

Gretchen

To unsubscribe from this group and stop receiving emails from it, send an email to plots-waterquality+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Amir Bernat]

Taking qualitative images with smartphones is definitely interesting but challenging, I'd be happy to know how are you planning on tackling this problem. There are some companies/apps which are doing this, three I liked from the medical world are Healthy.io for urine tests (which is really new), Billicam for neonatal jaundice detection (dating back to 2013ish) and F. Outlaw's recent publication (http://dl.acm.org/citation.cfm?id=3079488). 

Amir

[Gretchen Gehrke]

Hi Amir, 

The app works by taking a picture of a colorimetric formaldehyde detection badge. You take a "before" picture of the unexposed badge, then leave the badge open to the air for 24 hours, and then take an "after" picture of the badge. The app compares the before and after pictures, to determine the extent of reaction, and thus determines the formaldehyde concentration. 

The actual colorimetric comparison is not straightforward, as ambient lighting can create some real issues. Check out this interesting activity at the bottom of this page (under Visual Psychometrics) that demonstrates some of the complexity with lighting and background: https://www.morphtec.com/technology/. 

Thanks for the info about the other smartphone colorimetric tools. It'll be interesting to compare approaches! 

Best, 
Gretchen

--
Post to this group at plots-spectometry@googlegroups.com

 
Public Lab mailing lists (http://publiclab.org/lists) are great for discussion, but to get attribution, open source your work, and make it easy for others to find and cite your contributions, please publish your work at http://publiclab.org
---

You received this message because you are subscribed to the Google Groups "plots-spectrometry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plots-spectrometry+unsub...@googlegroups.com.

[Amir Bernat]

Thank you, Gretchen,

It is my experience that control of both illumination and imaging are required to be able to perform a quantitative analysis. IDK how sensitive are the stickers and how delicate is the color change presented, but keep in mind that setting of many camera parameters might be needed (exposure, ISO, WB, Gamma are probably the most important ones. Do you know is the visual change is the hue, saturation or value (using HSV jargon)?

There are thresholds for color accuracy in the LAB color space in general (Delta E <20), do you know what those are for this application?

Amir

Post to this group at plots-sp...@googlegroups.com

[Gretchen Gehrke]

Hi Amir, 

The colorimetric change is on the hue scale. I'm not sure if the value or saturation is expected to change -- that's a great question. I also don't know how the expected colors rate in terms of accuracy. If you ask these questions in the github repo (https://github.com/publiclab/SmART-Form), the app developers may be able to answer. 

Best, 

Gretchen

To unsubscribe from this group and stop receiving emails from it, send an email to plots-spectrometry+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
Post to this group at plots-spectometry@googlegroups.com

[Dave Stoft]

True, both a known source illumination and a color reference are required for absolute color measurements. However, the SafeAir badge analysis might use the Left:Right side colors in a differential measurement to avoid the color reference requirement. The badge colors have '0' Blue component and the Before:After RGB value differences are Left: 0:-32:0 and Right: +26:-49:0. It might be possible to base the measurement on color ratios for R/G and use any residual B as an error offset correction. Without a color reference, it seems likely the survey is looking to discover the 'generic' phone camera variance and thus the measurement accuracy limits.

Cheers,
Dave

[Amir Bernat]

Gretchen, thank you, will do that!
Dave, there are two problems (to begin with) just looking at the RGB ratios: the B values will be non-zero, and the ratios will change when the WB will change.

I haven't checked to see how does the delta hue change under different illumination or camera settings. Are you aware of any reference on that?

Amir

[Dave Stoft]

Amir,
Yes, color cast, due to the illumination source, is a real issue and a blue cast is often apparent in photos. That's why I said using a color reference was a requirement for absolute color measurements. For the SafeAir color (Before:Right) 200:115:0, adding a B component of 32 only decreases H by only 2. So, I was suggesting that, perhaps, they had developed correlations which would allow treating any B cast as an error term. Their 'public testing' seems oriented toward answering your question of how variable are the results from arbitrary uses of a variety of call phone cameras. If they intended to use a color reference, that information would be of less interest. Using a color reference is also more complicated -- which is why I guessed they were attempting to rely on RG ratios; and attempt to avoid major errors from the color cast issue. By measuring the 'before' values, for relative measurements, the colors of the first image do form an approximation of a color reference when you're only looking for color change -- even with a small amount of color cast.
Dave

[Amir Bernat]

Thank you for clarifying this Dave,

Would be a neat trick!

Amir