Extracting png files from website

If you read the entire page as HTML and then search through it for the png filenames and download those separately using the ImageFetcher it should work, although it kinda depends on how the png's are referenced in the page source. If it's <IMG> tags it should work but if there's a scripting language involved it might be trickier.

First step would be to inspect the page source though and see what you can find.

--edit--

Just looked at the page source myself, it doesn't look like there's any (useful) png's referenced in there so it looks like you're out of luck.

I think you want to scrape all the tiles from the tilelayer to recreate the map? First I need to say I think this is not a quick job so doing this would not be my first option. Requesting a copy from the dataset is a better alternative. But if you really want to it is doable.

Possible steps are:

• Understand how the URL is constructed.

https://spark-viewer.wivolo.com/tiles/render?dataset=u850.0108&fmt=png32&nocache=1545350950&index=16/63416/42528 

dataset is the layer

16/63416/42528 describes zoomlevel/x/y, where x and y are rows and columns

• Read and understand how the Bing/Google Tile Scheme relates to geographic coordinates.
• Determine zoomlevels and zoomlevels you need. I see a range from 5 to 16, you probably only need the deepest level (16). For developing and iterating through the data I would start with the highest level, 5. When it works as desired, convert to level 16.
• Determine which datasets you need and look them up. 4G is dataset=lte1800.0028, 3G = dataset=u850.0108, etc.
• Determine the min / max rows and columns you need.
• Generate the urls based on dataset / zoomlevel / tiles.
• Request the images using a HTTPCaller.
• Georeference the images.
• Merge the georeferenced images (RasterMosaicker).

I created something similar in the past for reading WMS. You might be able to reuse bits and pieces. See https://knowledge.safe.com/questions/1694/wmts-reading.html 

Yes I went down this route to but didn't get anywhere doing it this way. Thanks for the info though. This might work on other images I am trying to scrape from other websites.