[Req] Split and Combine a simple division

[DEN]

I would like the options that can be combined with a simple division.
It is optional, such as for example the attached image.

This image is just one example.
This feature is Option or Filter.

I also wrote in another thread earlier,

about out-of-memory data,
Write in to the virtual memory file in XnConvert folder.

[oops66]

Hello,
+1 I support

[cday]

There have been a number of requests for a Split function, at least.

Wouldn't this be Image |Split... ?

Maybe in XnViewMP first, so that it could be easily ported to XnConvert?

A start might be to split an image equally into left and right halves, or equally into top and bottom halves.

As one file would become two, thought needs to be given to the naming of the output files: one solution might be to add suffixes to the input file name such as -1 and -2, preferably user selectable so that the split images could be arranged in a desired sequence.

[Later evolution might be to split into four, as suggested by DEN, or even ultimately to an arbitrary number of arbitrarily defined areas: being able to inset crop boundaries might be useful to avoid edge shadow when splitting scans.]

cday.

[oops66]

Maybe in XnViewMP first, so that it could be easily ported to XnConvert?

...Yes

...Splitting image is very interesting with very big pictures (or maps), maybe the best and more efficient way is to cut the image like Slippy map tilenames - quadtiles (even for images, not only for geo-tagged maps)
http://wiki.openstreetmap.org/wiki/Slippy_map_tilenames

Code: Select all

This article describes the file naming conventions for the Slippy Map application.
    Tiles are 256 × 256 pixel PNG files
    Each zoom level is a directory, each column is a subdirectory, and each tile in that column is a file
    Filename(url) format is /zoom/x/y.png 

Code: Select all

 Derivation of tile names
    Reproject the coordinates to the Mercator projection (from EPSG:4326 to EPSG:3857):
        x = lon
        y = arsinh(lat) = log[tan(lat) + sec(lat)] 
        (lat and lon are in radians) 
    Transform range of x and y to 0 – 1 and shift origin to top left corner:
        x = [1 + (x / π)] / 2
        y = [1 − (y / π)] / 2 
    Calculate the number of tiles across the map, n, using 2zoom
    Multiply x and y by n. Round results down to give tilex and tiley. 

http://wiki.openstreetmap.org/wiki/QuadTiles
http://www.maptiler.org/

Code: Select all

QuadTiles are a geo-data storage/indexing strategy - it's more commonly referred to as hierarchical binning. The idea is to store a geo-database such that data for a specific location can be retrieved quickly, by dividing the data up by location, partitioning the world into tiles. Note that we use the word tiles a lot in OpenStreetMap, and in fact we're normally talking about tiled map images, that's not really what this page is about although a QuadTile data indexing scheme has some similarities with Slippy map tilenames. 

...But I think this project is too big for a single developer - Pierre

[cday]

Yes, but...

Image | Split... into equal left and right or top and bottom images would be a start...

A split into four might be then possible with successive left-right and top-bottom splits, if suitable output file name suffixes can be set.

cday.

[XnTriq]

There have been a number of requests for a Split function, at least.

• Split images into two halves
• Split image into two halves using crop command lossless JPG
• JPEG Lossless Join
• Split images?
• Cut A3 in half, save two A4?

[oops66]

... The goal of this functionality, is to be able to do this kind of panorama: (stitching + splitting) - lol
http://www.paris-26-gigapixels.com/index-en.html

Paris 26 Gigapixels is a stitching of 2346 single photos showing a very high-resolution panoramic view of the French capital (354159x75570 px). Dive in the image and visit Paris like never before!

---- In French ----------------------------
http://www.paris-26-gigapixels.com/index-fr.html
http://blog.paris-26-gigapixels.com/fr/?page_id=2
Paris 26 Gigapixels est un assemblage de 2346 photos individuelles représentant une vue panoramique de très haute résolution de la capitale (354159x75570 px). Plongez dans l'image et visitez Paris comme jamais auparavant !

[cday]

Split and Combine are separate functions although related: Split is clearly defined and has more support, Combine would benefit from being more clearly defined and might be better continued as a new thread.

A simple Image | Split... function as outlined above has good support and should be relatively easy to add?