00:00:49	Alex Styer (he/him):	Sorry if I missed this, but is the smoothing happening in 3x3 and 5x5 grids? And taking the average value of the grid?
00:01:52	Alex Styer (he/him):	Makes sense!
00:15:05	Sean Kearney (USDA-ARS):	both capture edges with high contrast between light/dark. h3 seems to highlight row-wise edges more where there is contrast between rows and the h4 seems to highlight column-wise edges.
00:20:50	Sean Kearney (USDA-ARS):	the light values are the same in both versions
00:21:13	Sean Kearney (USDA-ARS):	but now we can see the 'opposite' edges highlighted as dark values
00:22:47	Sean Kearney (USDA-ARS):	Are the Sobel filters also used to make hillshades, or is that a different algorithm altogether?
00:23:43	Sean Kearney (USDA-ARS):	ok - no problem!
00:25:43	25RS Serge Edme:	is it possible to make a composite of h3 and h4 and keep the best of the 2 on the new image?
00:27:12	Brian Stucky:	@Sean - Your question made me curious, so I investigated just a bit, and it looks like hillshading algorithms are based on simulating light/shadow areas of a surface when illuminated by a hypothetical point light source.  So quite a different approach, it would seem!
00:29:01	Sean Kearney (USDA-ARS):	Here's one approach to just take the max:
00:29:02	Sean Kearney (USDA-ARS):	plt.figure()
plt.imshow(np.max([abs(I_camera_h3), abs(I_camera_h4)], axis=(0)), cmap='gray')
plt.show()
00:29:56	Sean Kearney (USDA-ARS):	@Brian - that makes sense, thanks!!
00:32:03	Joe Kawash:	Are the Sobel filters used similarly to how OpenCV uses watershed transformation for identification of discrete "pieces" of an image?
00:33:31	Joe Kawash:	thanks!
00:36:46	Heather Jones:	Today's notebook is posted to the site to review.
00:37:39	Tavis Anderson:	Thank you very much!
00:37:58	Alex Styer (he/him):	Thanks Laura!
00:38:03	Daniel.Brabec:	Thanks > good info.
00:38:10	Sean Kearney (USDA-ARS):	Thanks! That was great. see you all tomorrow
00:38:12	Aaron Szczepanek:	Thanks laura and everyone!
00:38:14	25RS Serge Edme:	Many thanks!
00:38:15	Huihui Zhang:	Thank you!