command line softwares for 3D image processing in pool server of BL20XU

back

Pool server of BL20XU equips not only reconstruction programs of CT data but also several command line tools for image processing.
The tools are middle wares of slice softwares with C-shell scripts for easy operation of image processings those frequently applied for the analysis of CT data.

list of tools
srec_cuda - reconstruct and normalize CT images,
viewObj,- view 3D distribution of objects
viewhole,- view 3D distribution of holes
propObj,- analyze properties of objects
propHole,- analyze properties of holes

srec_cuda - reconstruct and normalize CT images (source code)

Usage: srec_cuda L1 C1 L2 L2 pixelsize bit [offset]
L1 : first slice number for reconstruction
C1 : rotation ceter for L1
L2 : last slice number for reconstruction
C2 : rotation ceter for L2
pixelsize : pixel size of detector
bit : bit per pixel for normalized images
offset(Optional) : offset angle for reconstraction

This program reconstruct CT images as a stack of tiff files with full range of LAC of each tiff file, and normalize the contrast of the images in the stack with a certain range of LAC. The range of the LAC and paramters used in the reconstruction are stored in rec.log. Full range images and normalized images are stored in L1 and L2 sub-directories, respectively.

viewObj - view 3D distribution of objects (source code)

Usage: viewObj dir lower upper [binning] [cluster-low cluster-high]
dir : directory name where tif files stored
lower upper : lower and upper thresholds of pixel value (gray value) for object
binning (Optional) : number of pixels to be reduced into 1 pixel. This parameter accelarates the processing for quick look, but spatial resolution becomes low.
cluster-low, cluster-high (Optional): range of order of cluster size to be displayed (1 gives the largest cluster and 65535 gives smallest clusters)

This program creats 3D images of objects included in the sample. Topview (same direction of slice image) and sideview (same direction of projection image at 0 degree) images are created.

sample images

viewHole - view 3D distribution of holes in objects (source code)

Usage: viewHole dir lower upper [binning] [cluster-low cluster-high]
dir : directory name where tif files stored
lower upper : lower and upper thresholds of pixel value (gray value) for object
binning (Optional) : number of pixels to be reduced into 1 pixel. This parameter accelarates the processing for quick look, but spatial resolution becomes low.
cluster-low, cluster-high (Optional): range of order of cluster size to be displayed (1 gives the largest cluster and 65535 gives smallest clusters)

This program is almost same with viewObj. Lower and upper threshold values indicate "object body", and holes inside the "object body" are viewed.

sample images

propObj - analyze properties of objects (source code)

Usage: propObj dir lower upper thick binning [cluster-low cluser-high] [lon lat] outdir
dir : directory name where tif files stored
lower, uppper : lower and upper threshold of pixel value (gray value) for objects in CT images
thick : number of pixels to be removed during grain identification analysis (between 1 to 10).
binning: number of pixels to be reduced into 1 pixel. This parameter accelarates the calculation for quick look, but spatial resolution becomes low.
cluster-low, cluster-high (Optional): range of order of cluster size to be analyzed and displayed (1 gives the largest cluster and 65535 gives smallest clusters)
lon, lat (Optional): angle of viewpoint of 3D images. (0 0) is topview (i.e. same direction of slice image) and (0 90) is sideview (i.e. same direction of projection image at 0 degree). Default is (-10 80)
outdir: directory name for output images. Colored images will be stored

This program creates color slice images and 3D imgages of objects having range of pixel value from lower to upper thresholds. This program also calculates triaxial ellipsoid approxemation of object shape, and creates a diagram for b/c and a/b (a < b < c, see detail in reference of sliceOF (link for slice)) and a 3D image of approxmated shape of objects. Stereographic projection images for the direction of three axes are also created using the result of triaxial ellipsoid approximation.

sample images

propHole - analyze properties of holes (source code)

Usage: propHole dir lower upper thick binning [cluster-low cluser-high] [lon lat] outdir
dir : directory name where tif files stored
lower, uppper : lower and upper threshold of pixel value (gray value) for objects in CT images
thick : number of pixels to be removed during grain identification analysis (between 1 to 10).
binning: number of pixels to be reduced into 1 pixel. This parameter accelarates the calculation for quick look, but spatial resolution becomes low.
cluster-low, cluster-high (Optional): range of order of hole size to be analyzed and displayed (1 gives the largest hole and 65535 gives smallest holes)
lon, lat (Optional): angle of viewpoint of 3D images. (0 0) is topview (i.e. same direction of slice image) and (0 90) is sideview (i.e. same direction of projection image at 0 degree). Default is (-10 80)
outdir: directory name for output images. Colored images will be stored

This program is almost same with propObj. This program calculates properties of holes inside the object(s). Lower and upper threshold values indicate "object body", and holes inside the "object body" are analyzed. Any other processes are same with propObj.

sample images

back