Support Website Contact Support Blog

noalph reflectance vs transparence reflectance

I have processed multi-spectral images for reflectance using Pix4D. I note two types of reflectance images created 

xx_desktop_noalpha_reflectance_xx

xx_desktop_transparent_reflectance_xx

what are differences between these two images?

If I want to convert the image as normalised reflectance values (range between 0.0000 - 1.0000), which data/index (noalpha or transparent) image that I should be using and where do I get the coefficients to convert image DN values to real numbers reflectance values?

thanks

Rajendra

Hey Rajendra, 

Are you using the the calibration target while you are processing? And what sensor are you using? If you are using Parrot Sequoia and MicaSense RedEdge you don’t need to normalize the data since it is being normalized in Pix4D. If you are using Slantrange you need to normalize the data. 

Hey Selim,

We used Parrot Sequoia and yes I noted values in transparent_reluctance recorded as real number. Just wondering if the incoming sensor need to calibrated before each flight?

Also, my one question was: 

what are differences between these two images?

xx_desktop_noalpha_reflectance_xx

xx_desktop_transparent_reflectance_xx

any document describing these products?

One more, we noted two key problems associated with Pis4D mosaic products:

  1. geo-referencing of RGB mosaic and individual band mosaic not matching

  2. data manipulation while creating mosaic (mainly in RGB mosaic) - I am more than happy to explain this with graphics, should you wish further information do not hesitate to call me +61 2 6764 5929 (Sydney time). I guess, Pix4d has mosaic editing tool, how can I access?

thanks

Rajendra

 

Hey Rajendra,

Before each flight Parrot Sequoia takes three sets of calibration images,  you don’t have to do anything else. And before start processing I would pick the calibration images that has the most most variance(no under or over exposed). I am not not %100 sure but judging by the name one of them has an alpha band (contains the no data) and the other one doesn’t have an alpha band. 

As for the geo-referencing are you using GCP’s ? The reason that RGB band and individual bands are not matching is due to the position of each sensor on Parrot Sequioa camera. If you are using GCP’s they should align with each other. And if you are not using GCP’s you can geo-reference them to each other or to a reference image in QGIS or ArcMap after you are done processing in Pix4D. 

You can use the Mosaic Editor tool to manipulate your mosaic, make sure you generate a DSM otherwise Mosaic Editor is not enabled and I believe you can only do it for RGB mosaics. 

Hi Rajendra,

The information Selim gave you is correct.

For what concerns the two TIFF outputs, the “noalpha_reflectance” gives a black color value to the no value data pixels. It will result in a .tiff file with black color where there is no information.
The “transparent_reflectance” does not display color information (it applies a transparent band) to the no value data pixels. It will result in a .tiff file where areas with no are not displayed as black color.

It is the same principle as for the different outputs of orthomosaic: https://support.pix4d.com/hc/en-us/articles/202558809

Regarding the other problems you are facing, please create separate requests to our Support team by filling this form

Hi Selim,
I used the Slantrange 4P+.

I would ask to you, in this case, how I can normalise the reflectance values.

Thank you.
Maria Angela