Export NDVI map

Hello

I aim to be able to count the trees (pine) within an area.
take the flight with an inspire1 + a RedEdge-M camera from Micasense.

I do the process, I calibrate the images and everything goes well up there.
when I want to export the data I export all the bands separately and I also export the NDVI Index, but when I take it to other analysis software such as Arcgis I see the monochromatic NDVI layer, I can not see it in color to start the classification process.

Please I need some guidance I am a novice in this.

Thank you.

Hi Rodrigo,

You will need to export the color index map: https://support.pix4d.com/hc/en-us/articles/202560539-How-to-generate-a-Colored-Index-Map-in-the-Index-Calculator. It will generate tif, jpg, png and also the color scale in a pdf file and will be stored in_. …\project_name\4_index \indices__\INDEX_NAME._ The name of the files will be projectName_rgba_ndvi.format.

Hello Momtanu

We were following what you explained to this gentleman some time ago.

The color index map in Pix4D is showing the NDVI values as they were when corrected using the spectrometer in field.

But when I export this color index map and reopen it in ArcGIS, I obtain an image with different RGB bands and not nir, green and red bands as called in Pix4D, and on top of that, they do not contain NDVI values but just the bands with values from 0 to 255.

How do I correct that?

Thank you

 

Hi Joseph,

The colored index map is exported as tif, png and jpg, but it is only for visualization (in case you want to add it in a presentation). It will be a RGB image. However, if you want to use it for further analysis in ArcGIS or QGIS, you will have to use the file generated by Pix4D in the output folder (for example, Sequoia_demo_dataset\4_index\indicesdvi\Sequoia_demo_dataset_index_ndvi.tif). It will show up as a monochromatic layer in QGIS in grayscale as it is a single band (NDVI having values from -1 to 1). You will have to colorize it in QGIS itself.

Hello 

Thank you for your answer.

We are doing a calibration using a spectrometer which allow us to correct the NDVI regular equation using the value of white and shades of grey measured with the spectrometer in field at the time of flying. As this correction is done once the orthomosaic is already created, I doubt that this correction is present on the orthomosaic present in the output folder, and I fear it will be the original orthomosaic uncorrected. Am I right?

When I export the colored index map as a tif, it is indeed in “RGB” (I don’t know if it is a regular RGB or if ArcGIS is calling the bands RGB but they would represent the real nir, red, green bands). I was wondering if the band values for each pixel would be corrected using the equation from Pix4D (as compared to the images in the output folder), and then, we can use the equation calculator in ArcGIS to calculate the NDVI value with the regular NDVI equation. I tried in ArcGIS to use the corrected equation provided by Pix4D following the use of the spectrometer directly on the original orthomosaic from the output folder, but it doesn’t work.

Hi Joseph,

If the images have not been radiometrically calibrated before and the images are the raw images from the camera, the orthomosaic produced by Pix4D would be uncorrected. 

The colored index map is just RGB with the values of red, green and blue for that colored map. I mean each pixel has a combination of red, green, blue to get that particular NDVI color of the pixel.

After getting the orthomosaic of each band, you can correct them with your equation. Are you not using the radiometric panel or the DLS sensor from Sequoia?

Hello, 

We are indeed using radiometric panel (panel with shades from white to dark greys) and use a spectrometer to measure the reflectance on this panel. We do that at the same time we are flying our tetracam.

When back to the llab, we create the orthomosaic on uncorrected NDVI raw images (and it was advised to correct the orthomosaic and not the raw images). Once we have the orthomosaic, we use the calibration tool, to enter the data from the spectrometer, create an NDVI equation corrected for the reflectance, and apply it to create the “corrected” NDVI orthomosaic.

The idea is now to export this corrected orthomosaic as geotiff, so I can open it on ArcGIS and proceed to the rest of my analysis.

Can you explain me the steps to do so?

Thanks 

Hello, 

We are indeed using radiometric panel (panel with shades from white to dark greys) and use a spectrometer to measure the reflectance on this panel. We do that at the same time we are flying our tetracam.

When back to the llab, we create the orthomosaic on uncorrected NDVI raw images (and it was advised to correct the orthomosaic and not the raw images). Once we have the orthomosaic, we use the calibration tool, to enter the data from the spectrometer, create an NDVI equation corrected for the reflectance, and apply it to create the “corrected” NDVI orthomosaic.

The idea is now to export this corrected orthomosaic as geotiff, so I can open it on ArcGIS and proceed to the rest of my analysis.

Can you explain me the steps to do so?

Thanks 

Hi Joseph, 

I hope I have understood this correctly :slight_smile: You are creating the NDVI equation along with radiometric calibration in some other tool and then using the index calculator to generate the index. You can export the index as geotiff and open it in ArcGIS. I have attached two screenshots. I have created an index named RG=R/G and exported it as geotiff.

 

Hello, 

So it is exactly what we want to do and globally what I was mentioning we did. That is the method we tried, but when I open the raster in ArcGIC, I obtain a RGB type of raster, with the 3 bands R, G, and B, instead of the NDVI values from -1 to +1 as expected (and present into the index map raster in Pix4D). I will retry again and let you posted.

thanks