Thermal expert Q1 versus Compact pro - tiny scale measures

[morningglory747]

Greetings,

I am currently undergoing a research project where I must acquire the temperature of a stigma. Stigmas are miniature components of a pistil (the reproductive organ of a flower) located in the middle of it. They are about a 0.5mm-1mm in diameter and 1-1.5mm in length. We have tried using thermocouples to measure the temperature, but these readings are not very precise and generate errors under solar radiation. We have concluded that thermal imagery is our best candidate and now we have to see which tool is the most optimal. Our budget is low so we decided to opt for thermal cameras that get attached to phones as they are not only cheap but also quite precise.

We are now debating on wether or not we should take the TE Q1 or the Seek Thermal Compact Pro. We believe the Compact Pro to be able to have a few squares on the stigma, but it seems like the TE Q1 could be even better, but I am hesitant to say as much, since the FOV of the Compact Pro is shorter (around 30 degrees while the TE Q1 is around 70). I have scrutinize many pages on this forum for answers, but I feel like asking is my best shot.

There is also the issue regarding image processing on the computer, which is something we would like to do. Apparently Seek Thermal Compact Pro has announce an update (quite sometime ago) that radiometric files will be available, but I cant seem to verify if they currently are or not. Is there anybody that can confirm this for me? On the other hand, TE Q1 has the csv files which contain temperature data for every grid point.
Note: We would do the image processing with a R package called Thermimage.

TL; DR:
Can I measure objects that are in the millimeters with the TE Q1 and is it better than the Compact Pro for these small objects?
Does the Compact Pro have any radiometric files that are available?

Thank you very much

Edit: Sentence structure

[Vipitis]

That is a thought one. The Thermal Expert is probably he better choice.

What is the time scale of your project? One or two weeks? Maybe you can rent a high end thermal camera and closeup lens for that time and expand a similar budget.

Lens will be difficult to focus close up. Ultrapurple did some thermal microscopy with a ThermApp and using two lenses front to front https://www.flickr.com/photos/ultrapurple/23896748304 - the Flickr gallery is wonderful and you will find some thermal images of flowers to judge size.

The ThermApp is more pricy and good lens will be as well.

I am not sure how good a PCB compares to a flower but it's likely that it's similar in size for tiny IC legs. I saw people using diy close ups lenses on Seek for PCB use.

Measuring fine temperature is going to be an issue, but you will always have relative temperature and when you can move a hot and cold point into frame to give you reference point you can do the calibration for a linear scaled image.

[morningglory747]

The scale of the project is about 10 days (thats how long the flowers stay). I was thinking about rent too as it does seem like a great idea, but I havent been able to find companies that do such a thing with their products. I will try to look more into it now since I kinda got fixated on the whole camera add-on thing.

Those are some wonderful pictures! The setup for some of them is pretty wack though and I don't think we're gonna get a chance to instal ourselves in such a way, but it doesn't seem like such fine level of detail is necessary (yet that remains to be seen).

The reference temperature is an important point. Maybe I could add some cold or warm surfaces next to (or behind) the object?

Maybe the TE Q1 plus will do the trick with its 13mm lens, but i'll see what I can do for the renting first.

Thanks!

[morningglory747]

I apologize for the double post, but I want to post an update and hear of your different thoughts on the matter.

I was able to find a rental FLIR E75 camera for about 390$ usd/week, its regular price is about 7k, so I suppose that it’s a good thermal camera?

When looking at the specs, I find it hard to see the difference between the FLIR E75 and the TE Q1, but maybe it’s just because I’m looking at the resolution? The FLIR E75 has a resolution of 320x240 while the TE Q1 is 384x288, so wouldn't the TE Q1 better even better?

From what I understand, the minimum focus distance, focal length and aspect/magnification ratio are also important elements to consider, but I fail to understand how these can help me out for reading the temperature data on small objects. My basic knowledge in photography is quite limited and I’m not particularly interested in detailed physical explanations, mostly just which numbers I should be looking for.

Here are the specs for the FLIR E75: https://www.flir.com/products/e75/
And for the TE models specs: https://www.thermalexpert.eu/english-1/product-overview/

Thank you

[Spirit532]

The TE Q1 is built down to a very low cost, with the optical path being entirely uncompensated, and the optical configuration being super simple, as well as the sensor being physically rather small compared to the E75.
The E75 is a very, very high end camera(that I believe actually has a 640x480 detector, crippled by firmware) with much bigger, brighter(faster) lenses and layers of calibration and filtering that provides cleaner, more accurate images.
When the Q1 states that its NETD is 50mK, you can safely expect a 40-60mK variation, and it would never dip below that unless heavily temperature stabilized and cooled(which it does not allow without serious hacking).
The E75 on the other hand, is actually lying about the noise of the sensor, but not in the way that you expect - it's adding noise on top of the raw image to cripple its performance on lower end models. On a clean image from an E75, without the noise added, you can safely assume you'll be seeing no more than 30-40mK of noise, and I've personally measured it to go as low as ~28mK on my T440bx(with a true 320x240 sensor and good optics, so comparable).

[Vipitis]

Flir also gives you flir radiometric images which you can use in flir tools or ResearchIR (if they still allow that 30 day trail it's perfect opportunity to do so).