Image package
The image package is part of the octave-forge project. See its homepage for the latest release.
Development
Follows an incomplete list of stuff missing in the image package to be matlab compatible. Bugs are not listed here, search and report them on the bug tracker instead.
Roadmap
The following are missing targets for the next minor release of the image package (version 2.8.0). Small bug fixes, build fixes, and regressions may still be released before in a patch release (Version 2.6.X).
- add ellipse properties to
regionprops
(see bug #44100) (done) - speed up
regionprops
for large images with many small objects (probably work around for loop with labeled == labelid) - support passing
bwconncomp
output toregionprops
(and then fixbwpropfilt
to use that instead of a labeled image) - ND support on
imfilter
- check if
normxcorr2
is Matlab compatible and investigate making use offftconvn
since it's likely to receive large input - investigate implementing the missing deconvolution functions (see patch #8571)
- anything else that is done or submitted before the above are done
Missing functions
- activecontour
- adapthisteq
- affine2d
- affine3d
- applycform
- axes2pix
- bwconvhull
- bwdistgeodesic
- bwlookup
- bwpack
- bwtraceboundary
- bwulterode
- bwunpack
- convmtx2
- corner
- cornermetric
- cpcorr
- cpselect
- cpstruct2pairs
- dct2
- dctmtx
- deconvblind
- deconvlucy
- deconvreg
- deconvwnr
- decorrstretch
- demosaic
- dicomanon
- dicomdisp
- dicomuid
- dpxread
- dpxinfo
- fan2para
- fanbeam
- fliptform
- freqz2
- fsamp2
- ftrans2
- fwind1
- fwind2
- gabor
- getimage
- getimagemodel
- getline
- getpts
- getrect
- gradientweight
- graycoprops
- grayconnected
- graydiffweight
- hdrread
- hdrwrite
- houghlines -- (see hough_line)
- houghpeaks
- iccfind
- iccread
- iccroot
- iccwrite
- idct2
- ifanbeam
- im2java2d
- ImageAdapter
- imageinfo
- imapplymatrix
- imapprox
- imbinarize
- imboxfilt
- imboxfilt3
- imcolormaptool
- imcontour
- imcontrast
- imdisplayrange
- imdistline
- imellipse
- imextendedmax
- imextendedmin
- imfindcircles
- imfreehand
- imfuse
- imgaborfilt
- imgaussfilt
- imgaussfilt3
- imgca
- imgcf
- imguidedfilter
- imhandles
- imhistmatch
- imhmax
- imhmin
- imimposemin
- imline
- immagbox
- immovie
- imoverview
- imoverviewpanel
- impixelinfo
- impixelinfoval
- impixelregionpanel
- impixelregion
- implay
- impoint
- impoly
- improfile
- imrect
- imref2d
- imref3d
- imregconfig
- imregcorr
- imregdemons
- imregister
- imregtform
- imroi
- imscrollpanel
- imsegfmm
- imseggeodesic
- imshowpair
- imtool
- imwarp
- integralBoxFilter
- integralBoxFilter3
- integralImage
- integralImage3
- interfileinfo
- interfileread
- iptaddcallback
- iptcheckhandle
- iptgetapi
- iptGetPointerBehavior
- iptgetpref
- ipticondir
- iptPointerManager
- iptprefs
- iptremovecallback
- iptSetPointerBehavior
- iptsetpref
- iptwindowalign
- isicc
- isrset
- makecform
- makeConstrainToRectFcn
- makehdr
- makeresampler
- multithresh
- nitfinfo
- nitfread
- openrset
- otsuthresh
- para2fan
- projective2d
- regionfill
- roifill
- roifilt2
- roipoly
- rsetwrite
- ssim
- tformarray
- tonemap
- truesize
- visboundaries
- viscircles
- warp
- whitepoint
- wiener2
- xyz2double
- xyz2uint16
Different functions
- with version 2014a, Matlab introduced a new function imtranslate. This function was part of Octave's package since 2002 but Matlab version is completely different. It needs to be rewritten for Matlab compatibility.
Missing options
- @strel
- missing SE decomposition for the diamond shape
- approximation argument for the ball and disk shapes are not implemented
- the translate method is not yet implemented
- blockproc
- The name-value input arguments have not yet been implemented.
- colfilt
- the option to split the image in smaller blocks before performing the im2col and filtering is ignored. Since that option only allows to use less memory and has no effect on the output is only important for sliding filtering with big blocks size.
- impixel
- interactive selection of points (less than 2 input arguments) is not yet implemented. Ability to create such UIs needs to be implemented in Octave core first.
- imcrop
- messing around with alternative coordinate systems (XData and YData) are not yet implemented.
- imwrap
- no supports for SmoothEdges option
- hough
- no RhoResolution option
Contributing
- It'd help to group the missing functions above, as per [1] (similar to what was done in the signal package wiki page); please keep the original alphabetical list intact.
Open Bugs
bug | bug title | status |
---|---|---|
#38085 | image package: fails to build with --enable-64 (entropyfilt) | Confirmed |
#38087 | image package: graycomatrix incompatibilities with matlab | None |
#38345 | image package: __spatial_filtering__() always returns class double | Confirmed |
#41674 | image package: temporary arrays in imresize are doubles | None |
#43218 | image package: imshear causes more signal spread than expected and inconsistent vertical offset | None |
#44396 | image package: bwmorph shrink does not preserve Euler number | Confirmed |
#44799 | image package: imrotate chrashes when using 'spline' as interpolation method | Confirmed |
#44831 | image package: imrotate - different dimensions of output matrix | Confirmed |
#45088 | image package: rangefilt requires domain and image to have equal number of dimensions | Patch Reviewed |
#46745 | image package: imresize default method (bicubic) Matlab incompatible for the border pixels | None |
#48529 | image package: compilation warnings under Mac OSX | None |
#50071 | OF image: bwfill(...,4) in Octave behaves like bwfill(...,8) in MATLAB and vice versa | None |
#50151 | image package: normxcorr2 is not compatible with Matlab | None |
#50188 | image package: poly2mask does incompatible rounding | None |
#50257 | image package: imtranslate is not Matlab compatible | Confirmed |
#50341 | Missing Function imrect in image package | None |
#50429 | image package: xyz2lab changed slightly in Matlab (and other color conversion functions) | None |
#50656 | image package: bwerode does not return right size when there's more dimensions on conn | None |
#50708 | image package: add demosaic.m | Patch Submitted |
#50732 | Image package: infinite error loop when assigning a strel (class) object to a vector | Confirmed |
#50783 | image package: imlincomb is not compatible with Matlab | none |
#50874 | bwdist call causes application crash | Confirmed |
Patches
- #8059: [new function for image package] ranktf.m
- #8060: [image package new function] whitepoint.m
- #8722: implement impixelinfo and impixelinfoval
- #8813: image package: Extended image I/O for Analyze 7.5 format to handle big-endian images
- #8866: image package: add metainfo.xml file for Gnome Software
- #8571: image package: three deconvolution routines (Wiener, RL, RLTV)
- #8824: added affine2d class
- #9235: image package: improve filtering functions (stdfilt, entropyfilt) and add tests (also medfilt2, ordfilt2, rangefilt)
- #9244: image package: regionprops.m - new properties ConvexHull, ConvexImage, ConvexArea, Solidity
Image Processing Functions (this section is under construction)
This is a list of Image Processing functions by topic, including missing functions, related bugs and patched, and some alternatives.
Note: This section is under construction so this list is still incomplete
Basic Read and Write Image Files
Function | Octave Package Version | Related Bugs and Patches | Alternatives | Notes |
---|---|---|---|---|
imfinfo | 2.6.1 | |||
imread | 2.6.1 | cv.imread from mexopencv |
in case of color image, the color order of Octave is RGB/RGBA while OpenCV color order is BGR/BGRA | |
imwrite | 2.6.1 | cv.imwrite from mexopencv |
in case of color image, the color order of Octave is RGB/RGBA while OpenCV color order is BGR/BGRA |
Type Conversion
Function | Octave Package Version | Related Bugs and Patches | Alternatives | Notes |
---|---|---|---|---|
adaptthresh | - | cv.adaptiveThreshold from [https://github.com/kyamagu/mexopencv mexopencv | ||
demosaic | - | #50708 | cv.demosaicing from mexopencv | |
gray2ind | 2.6.1 | |||
grayslice | 2.6.1 | |||
graythresh | 2.6.1 | Octave version has more methods | ||
im2double | 2.6.1 | |||
im2int16 | 2.6.1 | |||
im2java2d | - | |||
im2single | 2.6.1 | |||
im2uint16 | 2.6.1 | |||
im2uint8 | 2.6.1 | |||
imbinarize | - | imbinarize (I) is equivalent to im2bw(I, graythresh(I)), cv.threshold from mexopencv | ||
imcast | 2.6.1 | Octave Only | ||
imquantize | 2.6.1 | |||
ind2gray | 2.6.1 | |||
ind2rgb | 2.6.1 | |||
label2rgb | 2.6.1 | |||
mat2gray | 2.6.1 | |||
multithresh | - | |||
otsuthresh | - | graythresh | otsuthresh(hist) is equivalent to graythresh (hist) | |
rgb2gray | 2.6.1 | |||
rgb2ind | 2.6.1 |
Image Filtering
Function | Octave Package Version | Related Bugs and Patches | Alternatives | Notes |
---|---|---|---|---|
bwareafilt | 2.6.1 | |||
bwpropfilt | 2.6.1 | |||
convmtx2 | - | |||
entropyfilt | 2.6.1 | |||
fibermetric | - | |||
freqz2 | - | |||
fsamp2 | - | |||
fspecial | 2.6.1 | |||
ftrans2 | - | |||
fwind1 | - | |||
fwind2 | - | |||
gabor | - | |||
imboxfilt | - | |||
imboxfilt3 | - | |||
imfilter | 2.6.1 | |||
imgaborfilt | - | |||
imgaussfilt | - | |||
imgaussfilt3 | - | |||
imguidedfilter | - | |||
imsmooth | 2.6.1 | Octave only | ||
integralBoxFilter | - | |||
integralBoxFilter3 | - | |||
integralImage | - | |||
integralImage3 | - | |||
medfilt2 | 2.6.1 | |||
medfilt3 | - | |||
nlfilter | 2.6.1 | |||
normxcorr2 | 2.6.1 | #50151 | ||
ordfilt2 | 2.6.1 | |||
padarray | 2.6.1 | |||
rangefilt | 2.6.1 | #45088 | ||
stdfilt | 2.6.1 | |||
wiener2 | - |