Implement Normalizing Difference image function in core.py about aestheta

[kkmcgg]

Hey Jeeah!

I think you're understanding this right but just to be clear: you will not need to access the tile information directly to get the function working function. The function should be stand alone and accept any image, such as the google sat tiles as an argument:

So! To run the function you will need some image data as an input. You should be able to call the getTile() function to get some google sat images for example, and feed that into the new function. The getTile() function returns tiles as 3-D numpy arrays (rasters with height, width, and 'depth' - which represents the various bands or image colors).

So,

IMG_RGB = core.getTile([-63.5752,44.6488,2],source='google_sat')

then

ND = core.norm_diff(IMG_RGB, B1=1, B2=2)

So (I think this is your real question) to get access to the individual bands inside the function we can use the following type of numpy syntax:

In numpy this is called slicing . Slicing allows us to grab and modify certain parts of the numpy array by their index. So in terms of numpy, think of our images as 3-D cubes of pixels :

Now, yes unfortunately the order of our slices in our case, thinking about the image, is Y,X,Z. The magical : character slices all values of a given index. So if we place that for the Y and X axis of the array, our slice will return all Y and X values. Then if we set our third index Z as the index of our band number (for example band 1 is the index 0), our slice will return that band as a greyscale image.

So again, to grab a given band from an image represented as a numpy array, Arr, we use the following syntax:

[Jeeah]

I have a question. I know I need to make a function in core.py to do the math on the img bands but I am not sure how to access the bands from the google sat tiles.

[kkmcgg]

good work @Jeeah this function is working I'm going to close this issue