When e.g. WMTS or COGs are visualized, only the image data in the viewport (certain zoom and extent) is loaded. This makes the rendering time quite fast. But e.g. with low internet connection unloaded tiles are visible frequently when panning/zooming around.
To my understanding these image data (tiles) in the viewport have to be loaded entirely before they are rendered. I wonder if there are any possibilities to achieve a progressive rendering of tiles in the viewport, so that a low quality image is rendered with the first data bytes loaded and image quality enhances until the entire data is loaded? This would give a smoother visualization experience, since an entire viewport rendered in low quality is often more desired than if only one tile is rendered in high quality.
Recent updates to gdal added support for jpeg-xl compression which supports progressive rendering. A video showing the decoding process can be seen here: https://m.youtube.com/watch?v=UphN1_7nP8U&feature=emb_logo
Would progressive decoding be possible with geodata in theory or is it already implemented somewhere?