将大量图像拼接到一个千兆像素图像中 [英] Stitching large amounts of images together into one gigapixel image

问题描述

因此，我试图获取可供离线使用的我所在乡镇的卫星地图，并且我使用了通用地图下载器，可从bing的混合地图下载.一切正常，但是当我使用程序的内置图像拼接器时，我发现生成的位图图像太大，以至于它不是可用的位图. Photoshop和我可以接触到的所有图像查看器都无法查看(我猜测是因为位图的大小不应为6gb)

So I'm trying to get a sattelite map of my township that is available for offline use, and I used Universal Maps Downloader to download from bing's hybrid maps. That all worked fine, but when I used the program's built in image stitcher, I found that the resulting bitmap image was so large that it isn't a usable bitmap. Photoshop and all image viewers I can get my hands on won't view it (I'm guessing because bitmaps aren't supposed to be 6gb in size)

所以我想知道是否A.)是否有任何程序可以将一吨(〜45,000)较小的图像拼接/查看为一个大图像，或者B.)是否有任何可以处理此尺寸图像的文件格式?比位图好一点，我可以用来将较小的图像拼接成该格式的任何程序(UMD内置的拼接器只对位图有效)

So I'm wondering if A.) there are any programs to stitch/view a ton (~45,000) of smaller images as one large image, or B.) any file formats that would handle an image of this size a little bit better than bitmap, and any programs that I could use to stitch the smaller images into that format (the built in stitcher in UMD only does bitmap)

推荐答案

文件格式:GeoTIFF.或几乎任何格式都可以，只要您平铺即可.

File formats: GeoTIFF. Or almost any format will do as long as you tile.

程序:一旦您达到RAM限制并开始交换，如果您坚持将其保持在一个块中，那么任何操作都将不起作用.

Programs: Once you hit the RAM limit and start swapping nothing is going to work if you insist on keeping it in one piece.

正确的方法是平铺并跟踪查看器中的缩放级别和视口位置.用谷歌搜索平铺图像查看器".

The correct approach is to tile and keep track of zoom level and viewport location in the viewer. Do a google search for "tiled image viewer".

请注意，如果您使用有损压缩的图像格式(例如JPEG)，则最终可能会看到可见的图块边缘，因为边缘任一侧上的压缩都会产生不同的伪像.在这种情况下，通常有效的技巧是在较大的画布中渲染图块，然后在书写之前将其裁剪为图块大小.

Note that if you use an image format with lossy compression (e.g. JPEG) you may end up with visible tile edges because the compression on either side of the edge will yield different artifacts. A trick that usually works in that case is to render the tile in a larger canvas, then crop to tile size before writing.

这篇关于将大量图像拼接到一个千兆像素图像中的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！