Tiled Data Reconstruction and Correction

Tiled Data Reconstruction and Correction

ByAbhishek Singh

Rajhesh Babu, Bangalore RevannaAnd Joel Solomon, Bula

ContentsIntroductionImageJImage StitchingPlugin UsagesFuture WorkConclusionQueries?

IntroductionThere is an increasing demand to image large

biological specimen at high resolution, typically those specimen which do not fit in the field of view of the microscope.

Biologists use Confocal, Bright field or Electron microscopes equipped with motorized stages to stepwise image large areas using high magnifications.

The acquired image tiles have to be combined into one final output image by a process usually referred to as Image Stitching.

About ImageJImageJ is public domain, Java based Image

processing program that was designed with an open architecture that provides extensibility via Java plug-ins and recordable macros.

User-written plug-ins make it possible to solve many image processing and analysis problems.

 It can read many image formats including TIFF,GIF,PNG,JPEG as well as raw formats.

ImageJ supports image stacks, a series of images that share a single window, and it is multithreaded, so time-consuming operations can be performed in parallel on multi-CPU hardware.

Image StitchingImage Stitching is the process of the

combining multiple images to form a single high resolution image.

Image stitching in image j has 2 plugins Pairwise Stitching – The input is only 2 images

at a time.Grid Stitching – The input is an arbitrary

collection or grid of images.

Input tiff format file

Pairwise StitchingIt queries the two images that you intend to

stitch. The images can contain rectangular ROIs which limit the search to those areas, however the full images can be stitched.

Once you have selected the two images to be stitched, it will show the actual Pairwise stitching dialog.

The dialog will depend completely on the dimensions of the images selected, and you have to note that the RGB input images will be converted to 8bit composite images.

ProcedurePairwise Image stitching : input two imagesLinear Blending : Smooth adjustment Average : Compute average intensity Max IntensityMin Intensity Overlay into composite : Channel separation Computation parameters : Save memory Thresholding

Dialog of Pairwise Stitching

Example of Pair-wise stitching

Grid/Collection StitchingThis plugin is able to stitch an arbitary collection or

grid of images whether it is 2d-5d images, as long all images are of same type.

In contrast to Pairwise stitching, this plugin will load and potentially save the images from/to the harddisk.

It’s always better to give the approrimate values of the tiles of the tiles, this will reduce the computational effort of the Grid stitching plugin.

If the layout of the tiles is not known, then the Stitching will figure out on it’s own, but this takes time.

The first dialog queries the type image collection or image grid that you want to assemble.

For each major type, there are many sub-types you can choose from. The various options are :

Grid : row by rowGrid : column by columnGrid : snake by rowsGrid : snake by columnsFilename defined positionsUnknown positionsPositions from File

First dialog and second dialog

Code Snippet

Input

Sample Slices for stitching

Two images stitched

Output

Four images stitched

Output

4 slices stitched in a column

4 slices stitched in a row

Brightness Control Using Slider

Brightness Control

Libraries for plugins

Main classes and plugins

Conclusion and future workWe have to have to work on alignment and

scaling. “Out of memory” problem.We have implemented the conversion of .oib

image format to .tiff using bio-format reader.How to adjust the correct brightness in the

stitched images.Also work on image blending and

overlapping .

Referenceshttp://drosophila.biology.kent.edu/users/rclem

ent/resources/Jan30-2014.pdfhttp://fiji.sc/Image_Stitchinghttp://www.cs.bath.ac.uk/brown/autostitch/au

tostitch.htmlhttp://bioinformatics.oxfordjournals.org/conte

nt/25/11/1463.full.pdf

Thank you!!