img from analysis functions empty

Really nice job with the VIS tutorial! 1.5 things that stuck out to me: at the very end, for the analysis, the code posted doesn’t actually run in a jupyter cell because 1. indentation 2. args does not exist.

I removed those problems but then found that analyze_object(), analyze_bound(), analyze_color() all return empty arrays for the images. Looks like you only output the image if area and filename are defined. any reason for this? I realized pcv.params.debug = 'plot' also shows the plots but it be nice to have this flexibility to see the plots separately. As presented one would expect the images be output, even if filename (in my case) are not defined. If area is not defined you might actually consider a warning message.

[nfahlgren]

It depends on your preference, but what will happen functionally is this. Let’s say I put filename=test.png. pcv.analyze_object will output two images, test_shapes.jpg and test_mask.jpg, into the current working directory. If you run this analysis on multiple images this is probably not desirable because all the output images will have the same name. So as you said, I would generally use the filename variable that is output by the pcv.readimage function because you can then link the output image names to the input image name. The same goes for the output directory. You can either code this directly into your script or you can use an input variable to set the output directory. In your case you could set outfile = "C:\Users\Dominik Schneider\Downloads\" + filename to put all the output files in Downloads with the original image name plus the new suffix.

[nfahlgren]

@dschneiderch, no worries, this is awesome! In order to make the function available after import you have to add it to the __init__.py file one directory up: https://github.com/dschneiderch/plantcv/blob/readbayer/plantcv/plantcv/__init__.py