SubImage¶

MLModule¶

genre	`Subimage`
authors	`Florian Link`, `Tobias Boskamp`
package	`MeVisLab/Standard`
dll	`MLBasicModules`
definition	mlBasicModules.def
see also	`Magnify`, `VecBoxConvert`, `DecomposeVector3`, `GetVoxelValue`
keywords	`crop`, `box`

Purpose¶

The module SubImage extracts subimages from its input image.

A SubImage module can also create a region larger than the input image.

Details¶

Some details on the Mode 'WorldStartEndImageAxisAligned':

Note that this mode does NOT interpret the world positions as corners of a box that is world-axis aligned (e.g. like a CSO bounding box), but image-axis aligned (thus the long title).
Every voxel touched by more than epsilon for any epsilon>0 by the so-defined box will be included in the output image. Thus, a world box is most robustly defined by providing voxel centers.
A special case is the one where for the given world start/end positions w1 and w2, w1[i] == w2[i] for any dimension i. In this case, the extent of that dimension will be zero rather than 1 despite condition 2 (by convention, such a box is not empty but of extent 1).

Tips¶

The update of the output image of this module depends on the Apply/Apply Box buttons and the Auto Apply/ Auto Apply Box checkboxes.

If no auto apply is set, the SubImage module still does an explicit apply on each input image change. This means that the output image of the module is recalculated from the module's coordinate fields whenever the input image changes.

Windows¶

Default Panel¶

../../../Modules/ML/MLBasicModules/mhelp/Images/Screenshots/SubImage._default.png

Input Fields¶

input0¶

name: input0, type: Image¶

Output Fields¶

output0¶

name: output0, type: Image¶

Parameter Fields¶

Field Index¶

`Always clamp to input image region`: `Bool`	`Full Size`: `Trigger`	`U (u)`: `Double`
`Apply`: `Trigger`	`Mode`: `Enum`	`U (modifyU)`: `Bool`
`Apply Box`: `Trigger`	`Sc`: `Double`	`X (x)`: `Double`
`Auto Apply`: `Bool`	`St`: `Double`	`X (modifyX)`: `Bool`
`Auto Apply Box`: `Bool`	`Su`: `Double`	`Y (y)`: `Double`
`Auto-correct for negative subimage extents`: `Bool`	`Sx`: `Double`	`Y (modifyY)`: `Bool`
`Box Input`: `String`	`Sy`: `Double`	`Z (z)`: `Double`
`C (c)`: `Double`	`Sz`: `Double`	`Z (modifyZ)`: `Bool`
`C (modifyC)`: `Bool`	`T (t)`: `Double`
`Fill Value`: `Double`	`T (modifyT)`: `Bool`

Visible Fields¶

Box Input¶

name: boxInput, type: String, default: 0 0 0 0 0 0 -1 -1 -1 -1 -1 -1¶: Sets a box string whose first 6 components represent one corner point and are used to fill the 'start' fields, and the second 6 the other corner point or the boxes extent (depending on Mode), filling the 'size/end' fields.

Apply Box¶

name: applyBox, type: Trigger, deprecated name: updateFromBoxInput¶: If pressed, the currently specified box is used.

Auto Apply Box¶

name: autoApplyBox, type: Bool, default: FALSE, deprecated name: autoUpdateFromBoxInput¶: If checked, the currently specified is used on any field change.

Auto-correct for negative subimage extents¶

name: correctForNegativeSubImageExtents, type: Bool, default: FALSE¶

If disabled (default), any input that results in an output box with negative extents (v1[i] > v2[i] for any component i) will result in an invalidated output image. If enabled, the components of the output subimage box are sorted so that v1 <= v2 in all components.

Note that empty boxes can still occur for the voxel modes if the size is exactly 0, or if Always clamp to input image region is enabled and the box lies entirely outside of the input image.

Always clamp to input image region¶

name: enableBoxClamping, type: Bool, default: FALSE¶: If enabled, the input image can only be cropped, but never extended. Thus, the Fill Value has no effect in this case. Default: disabled.

X (x)¶

name: x, type: Double, default: 0¶: Sets the start x-coordinate of the subimage.

Sx¶

name: sx, type: Double, default: 32¶: Sets the size or end x-coordinate of the subimage.

X (modifyX)¶

name: modifyX, type: Bool, default: TRUE¶: If checked, the x-extent can be modified.

Y (y)¶

name: y, type: Double, default: 0¶: Sets the start y-coordinate of the subimage.

Sy¶

name: sy, type: Double, default: 32¶: Sets the size or end y-coordinate of the subimage.

Y (modifyY)¶

name: modifyY, type: Bool, default: TRUE¶: If checked, the y-extent can be modified.

Z (z)¶

name: z, type: Double, default: 0¶: Sets the start z-coordinate of the subimage.

Sz¶

name: sz, type: Double, default: 32¶: Sets the size or end z-coordinate of the subimage.

Z (modifyZ)¶

name: modifyZ, type: Bool, default: TRUE¶: If checked, the z-extent can be modified.

C (c)¶

name: c, type: Double, default: 0¶: Sets the start c-coordinate of the subimage.

Sc¶

name: sc, type: Double, default: 1¶: Sets the size or end c-coordinate of the subimage.

C (modifyC)¶

name: modifyC, type: Bool, default: FALSE¶: If checked, the c-extent can be modified.

T (t)¶

name: t, type: Double, default: 0¶: Sets the start t-coordinate of the subimage.

St¶

name: st, type: Double, default: 1¶: Sets the size or end t-coordinate of the subimage.

T (modifyT)¶

name: modifyT, type: Bool, default: TRUE¶: If checked, the t-extent can be modified.

U (u)¶

name: u, type: Double, default: 0¶: Sets the start u-coordinate of the subimage.

Su¶

name: su, type: Double, default: 1¶: Sets the size or end u-coordinate of the subimage.

U (modifyU)¶

name: modifyU, type: Bool, default: FALSE¶: If checked, the u-extent can be modified.

Fill Value¶

name: fillValue, type: Double, default: 0¶: Sets a fill value that is used for all empty regions in case the output image extends to regions not covered by the input image.

Mode¶

name: mode, type: Enum, default: VoxelStartSize¶: Defines the mode in that the size of the subimage is specified.

Values:

Title	Name	Description
Voxel Start & Size	VoxelStartSize	A starting point and a size in voxel coordinates need to be specified.
Voxel Start & End	VoxelStartEnd	A starting and an ending point in voxel coordinates need to be specified.
World Start & End (Image Axis Aligned)	WorldStartEndImageAxisAligned	Like VoxelStartEnd mode, but the start and end coordinate are to be given in world coordinates. Note that this mode does NOT interpret the world positions as corners of a box that is world-axis aligned (e.g. like a CSO bounding box), but image-axis aligned (thus the long title). Also, every voxel touched by more than epsilon for any epsilon>0 by the so-defined box will be included in the output image, so such a world box is most robustly defined by providing voxel centers.
Voxel Center & Width	VoxelCenterWidth	A middle point and size in voxel coordinates.

Auto Apply¶

name: autoApply, type: Bool, default: FALSE¶: If checked, the module computes anew on any field change.

Apply¶

name: apply, type: Trigger¶: If pressed, the module computes anew.

Full Size¶

name: fullSize, type: Trigger, deprecated name: getProperties¶: If pressed, all fields are set to the full values of the input image.

Navigation