EEVblog Electronics Community Forum

Products => Thermal Imaging => Topic started by: alex871 on January 24, 2020, 07:09:44 pm

Title: Software radiometric for elaboration immage HT-301
Post by: alex871 on January 24, 2020, 07:09:44 pm
hi all, these are the radiometric files of the 301HT, would anyone be able to find software that can view them and process the images?

https://www.dropbox.com/sh/fk5qjd9tbstzvsw/AABFNBUI-TWGRwsJtxeNlqF8a?dl=0 (https://www.dropbox.com/sh/fk5qjd9tbstzvsw/AABFNBUI-TWGRwsJtxeNlqF8a?dl=0)

 :)
Title: Re: Software radiometric for elaboration immage HT-301
Post by: JimM on January 24, 2020, 10:36:04 pm
Assume by elaboration you mean calculate temperatures.
The three files in the dropbox are: PNG file, JSON file and txt file.
The PNG file is just an image file with no metadata.
The JSON file has metadata, but not enough to calculate temperatures. Namely:
Code: [Select]
{"Imager info":"ThermViewer","Date + Time":"23/01/2020 00:02:36","Application Version":"2.0.19","Temperature scale maximum value":"32,2","Temperature scale minimum value":"29,1","Emissivity":"0.95","File save path":"/storage/emulated/0/ThermViewer/Export/20200123_000236/","Device Name":"T3","Device Serial number":"IJ1081","Device Firmware number":0,"Device Hardware number":0,"Resolution":"384X288","Reflected temperature":20,"Temperature compensation":0.0,"Temperature scale truncation":0.25,"Spot reading type":"Center","Spot reading temperature":"31,0"}

The txt file seems to be the raw data.
So there doesn't seem to be enough here to process the image.
Title: Re: Software radiometric for elaboration immage HT-301
Post by: alex871 on January 24, 2020, 11:13:33 pm
All temperatures are in the txt file, there are 110,592 values equal to the product of 384 x 288  ::)
Title: Re: Software radiometric for elaboration immage HT-301
Post by: IwuzBornanerd on January 24, 2020, 11:22:31 pm
The text file is a list of integers separated by blank spaces.

I used ghex to replace the blanks with newline character & read the file into my Seek manipulator (after modifying it for the new image size) and got the attached image.  I don't know if the temperatures are correct but they do look a bit high, and even at that I find it very surprising given how I convert the file values to temperature without knowing what the values actually mean.

At any rate, it should be simple enough for a programmer to make images out of the files & get the correct temperatures.  ;)

[attach=1]
Title: Re: Software radiometric for elaboration immage HT-301
Post by: JimM on January 24, 2020, 11:38:42 pm
Yeah, the txt file has 384x288 4-digit integers in it. But, are those the raw values or can they be interpreted as temperatures by placing a decimal point after the first two digits? Would be interesting to take a picture of something with negative temperatures and see what's in the txt file.
Title: Re: Software radiometric for elaboration immage HT-301
Post by: alex871 on January 25, 2020, 12:32:24 am
Quote from: IwuzBornanerd on January 24, 2020, 11:22:31 pm
The text file is a list of integers separated by blank spaces.

I used ghex to replace the blanks with newline character & read the file into my Seek manipulator (after modifying it for the new image size) and got the attached image.  I don't know if the temperatures are correct but they do look a bit high, and even at that I find it very surprising given how I convert the file values to temperature without knowing what the values actually mean.

At any rate, it should be simple enough for a programmer to make images out of the files & get the correct temperatures.  ;)

(Attachment Link)

Great!  :-+ :clap: can you explain me how you did it? It's difficult ?
Title: Re: Software radiometric for elaboration immage HT-301
Post by: IwuzBornanerd on January 25, 2020, 03:18:09 am
I'm not sure how much of "how I did it" you want so I'll start with the simple part.

I read the values into the array in my program that I call "tamfx22" for "Temperature Above Minus 40 times 22".  Minus 40 because that is Seek's lowest temperature; times 22 because the length of Seek's lookup table is about 22 values per degree Fahrenheit (closer to 40 per C).  For Seeks I pick the tamf value out of a lookup table based on the processed image data for each pixel.  For your image I just had the program read in the values & calculate the temperature based on the same scale, so a value of zero would have been -40C & then about 40.4 per degree C above that.  So whatever the number in the file was, the temperature shown would have been X/40.4 - 40.  Pulling the file into a spreadsheet I find the min & max values as 1816 & 3195 which gives temperatures of 4.9C & 39.0C, which are pretty close to what the image shows.

Oddly, though, my image is mirrored from yours so the pixel order may not be left to right top to bottom as I assume.

The json file says minimum 29.1 & max 32.2.  Is that right?  Hot floor??  If the files were saved by Thermviewer, isn't there some documentation to tell you what the data is?  Or maybe you can just ask Jinhua Ye.

Now getting the numbers onto the screen is the hard part...