Does Pix4D have any issue creating orthomosaics of thermal images with and without radiometric data? I would guess it can do an accurate job based on the RGB content just like any photo?
If the data is properly geo referenced, Pix4D should be able to geo-ortho-rectified image data. Was your project unable to mosaic at all?
Hi Jeff,
The orthomosaics are a 2D maps, with each point containing X, Y and a color information. They are therefore produced from RGB images. Concerning thermal images (with and without radiometric calibration), you should not use an orthomosaic but rather a reflectance map (physically speaking an emmitance map). On orthomosaic, we do color balancing whereas on reflectance map you have the raw value of the pixels.
We do recommend to use radiometric data since this would give you the temperature, thus using : .rjpeg or grayscale .tiff. The .jpg files would contain only a visual representation of the temperature. Further on the .jpg are coloured-mapped so then a pixel may have a different colour on different images which would again not represent the emissivity .
For more information on how to process thermal datasets in Pix4Dmapper, please refer to the following article: https://support.pix4d.com/hc/en-us/articles/360000173463
In case you would be interested in merging datasets with both thermal and RGB imagery: information from both thermal and RGB datasets cannot be used to generate a single orthomosaic. We recommend to create two separate projects (one RGB and one Thermal), and follow the instructions described in the link above.
Hope this helps,
Is there a way to process thermal images that have no geo data? I started flying a thermal camera on a homebuilt drone before I knew how to add the geo tags. I have GCPs that I can find in the jpegs but I can’t figure out how to go about doing this with the .tiffs
Hello @mkummel,
You can process the thermal images without geo data. To mark the GCPs on the .tiff images, you should first process step 1. I would suggest you to process step 1 first and then only mark the GCPs using the rayCloud.
Thank you for your response. I guess my question is that I can not visually see anything in the .tiffs. The greyscale visualization is just one shade. Is there a way to adjust that so I can see a gradient in the images?
How I make thermal ortho form Mavic 2 Enterprises Advance drone photo.