Pix4D Mapper Index Calculator
Hello! I'm Marc Degroiseiller from technical support here at Pix4D and today
I'm talking to you about the index calculator.
The index calculator is a tab in Pix4Dmapper that will give you a streamlined workflow to get all of the outputs that you're looking for in a precision agriculture project. And what do I mean by a precision agriculture project?
Well, it's a project where the main objective is to compute the reflectance of the different features in your in your field and so the first output that you're looking for is the reflectance map and you will all using the index calculator you will also obtain an index map, such as the ndvi map and then it will be possible to colorize this map
colorizing this map will make the data analysis easier.
Another output that you can get is the application file so your field is going to be segmented into different regions according to the index values and and there's going to be a shapefile that's going to be generated that will contain as an added
information: the amount of resources that should be given to the different regions
and this file can be put into a smart tractor if it accepts this kind of file.
So it makes the whole process quite automatized.
In many of these precision agriculture projects you're going to be using a radiometric calibration target while it's not mandatory it does allow to compare data sets taken
in different days or under different conditions and as such, many users and
have this calibration target and use it for their projects. And let me give you a few tips about how to bestuse this calibration target first of all it should be flat on the ground and not tilted.
Also while you're taking the picture, of the calibration target it's important
not to cast a shadow on the target and for the region with the known
reflectance to not be overexposed or underexposed because in this case and
the data is not usable so that means that very often it's a good idea to take
at least a few set of images to make sure that at least one set is good and
can be used for radiometric calibration this data should be inputted in the
mapper before you start the third step of processing so let's see in the
software how you can give this data to the mapper, to get ready radiometrically precise results let's see in the mapper how to input the calibration target values so as you can see here I haven't started the processing so i can go in the processing options...
and I go in: "3. DSM Orthomosaic and Index" and I go in the index calculator tab
you can see here that the different sensors have not been calibrated this is
a Sequoia project so i'm going to click on calibrate and I'm going to find the
appropriate image so you see here this is the green sensor so I browse to where
my images are and i'm going to select the green image you can see it here
going to draw and rectangle around where in the in the calibration region here
with the known reflectance, I do a right click to finish the shape now you see
it's important that this this shape that i drew is completely inside the region
and the value here I should know it they should be provided by the manufacturer
and I know that it's 192 for my particular target this number is going to differ from target to target and I click calibrate and now it's been done so that's how you do it i would repeat the same procedure for all the different sensors and that's how you input the radiometric calibration target information in the mapper. One thing to
note is that for multi-spec projects as well as for sequoia projects that are
processed with version 2.2 or later and with the firmware version 1.1 this is
done automatically so when the images of the regimental calibration target are
recognized and all this information is is put in the mapper automatically
but if you use a different target or if you use an older version of the fair
more on Sequoia for example you should go through these steps as well.
We can see now on the screen, the index calculator window and what's really nice
about this is that the workflow is quite linear you go from one step to the next
and this is how you're going to obtain your input the first step is the generation of the
reflectance map Now this reflectance map is really the core ingredient in the precision agriculture processing workflow and so how is the mapper going to obtain this reflectance map well it's going to compute a reflectance value for every
pixel and it's going to take into account many different variables such as
the camera position the the angle between the camera and the feature and
if the information has been provided in the EXIF file it's going to correct for
dark current, vignetting and other e ffects so really there are a lot of variables
and that are accounted for and this is the most computationally intensive step
this is where we really leverage the Pix4Dmapper engine and we use
the 3D reconstruction to be able to compute accurate reflectance values
Once the reflectance maps have been generated, you can produce an index map.
Index maps you can think of NDVI, so what does it do it takes into
account the different information from the reflectance maps and it puts
them together using a mathematical formula in order to make the analysis easier
so in the case of NDVI, you want to be able to look at this
NDVI map and figure out what is the health of of your plant in the different regions
Once the index maps had has been generated you can colorize it again with
the objective to make data analysis easier . And once you have these these
regions, depending on the index value you can assign to each region a certain amount of resource that should be spread over this this region so this resource it could be fertilizer but it could also be water for example and so you say well in this region which has NDVI value between 0.5 and 0.7, I would like to apply a certain amount of resource so you really segment your field according to the index values and then you're going to be able to export this information into a shapefile and
the shapefile you put it in your smart tractor if it accepts this kind of input
and it's also possible to export the colored index map if you need to share
it so it's going to be exported as a JPG file a TIFF file or a KML file for Google Maps
That's all I wanted to say about the index calculator today, I hope that what I've said is going to allow you to get the most out of Pix4Dmapper and i'll see you in the next video.