Write tiles in local machine

[jzphlp]

Hi there, is there an implementation with your that allow to tile a big raster file and save each tile locally? Big thanks in advance

[vincentsarago]

@jzphlp no I'm not aware of any tools for this based on rio-tiler. Mostly because gdal2tiles https://gdal.org/programs/gdal2tiles.html#gdal2tiles is perfect for this.

[akosmaroy]

I was also wondering if there is a tool or a script that generates a complete tile set using rio-tiler?

I'm aware of gdal2tiles, but it has it's limitations, for example with 4 band color GeoTIFFs it's not really working well in my experience.

[vincentsarago]

the issue is that such a tool can be easy to write but will fit only few cases (e.g. what happens if your file is not of type Uint8 ?), which is why I didn't want to write one in the first place.

here is a quick sketch of the code that could create tiff tiles for a specific file

import pathlib
from rio_tiler.io import Reader

with Reader(input) as src:
    w, s, e, n = src.geographic_bounds
    for zoom in range(minzoom, maxzoom + 1):

        pathlib.Path(f"tiles/{zoom}").mkdir(parents=True, exist_ok=True)


        ul_tile = src.tms.tile(w, n, zoom, truncate=True)
        lr_tile = src.tms.tile(e, s, zoom, truncate=True)

        minmax = self.tms.minmax(zoom)
        extremas = {
            "x": {
                "min": max(ul_tile.x, minmax["x"]["min"]),
                "max": min(lr_tile.x, minmax["x"]["max"]),
            },
            "y": {
                "min": max(ul_tile.y, minmax["y"]["min"]),
                "max": min(lr_tile.y, minmax["y"]["max"]),
            },
        }

        for x in range(extremas["x"]["min"], extremas["x"]["max"]):
            pathlib.Path(f"tiles/{zoom}/{x}").mkdir(parents=True, exist_ok=True)

            for y in range(extremas["y"]["min"], extremas["y"]["max"]):
                with open("tiles/{zoom}/{x}/{y}.tif") as f:
                    f.write(src.tile(x, y, zoom).render(img_format="GTiff", add_mask=True, compress="DEFLATE"))

[akosmaroy]

thanks, I slightly modofied it, works great:

import pathlib
from rio_tiler.io import Reader

def render_tiles(inputfile, minzoom, maxzoom, outputdir):
	print("Rendering {} to {}".format(inputfile, outputdir))
	with Reader(inputfile) as src:
		w, s, e, n = src.geographic_bounds
		for zoom in range(minzoom, maxzoom + 1):

			pathlib.Path(f"{outputdir}/{zoom}").mkdir(parents=True, exist_ok=True)


			ul_tile = src.tms.tile(w, n, zoom, truncate=True)
			lr_tile = src.tms.tile(e, s, zoom, truncate=True)

			minmax = src.tms.minmax(zoom)
			extremas = {
				"x": {
				"min": max(ul_tile.x, minmax["x"]["min"]),
				"max": min(lr_tile.x, minmax["x"]["max"]),
				},
				"y": {
				"min": max(ul_tile.y, minmax["y"]["min"]),
				"max": min(lr_tile.y, minmax["y"]["max"]),
				},
			}

			for x in range(extremas["x"]["min"], extremas["x"]["max"] + 1):
				pathlib.Path(f"{outputdir}/{zoom}/{x}").mkdir(parents=True, exist_ok=True)

				for y in range(extremas["y"]["min"], extremas["y"]["max"] + 1):
					print("\rzoom {}, tile {},{}".format(zoom, x, y), end='\r')
					with open(f"{outputdir}/{zoom}/{x}/{y}.png", "wb") as f:
						f.write(src.tile(x, y, zoom).render(img_format="PNG", add_mask=True, compress="DEFLATE"))