NIH - IMAGE MACROS FOR COCCOLITH RESEARCH

        WARNINGS:
        1. These macros are research tools not commercial software, we try to make them robust and easy to use but...
        2. This set of WWW pages to provide some documentation, but they are not a comprehensive guide. Please do contact us if you need help.

        Digital image capture provides a range of opportunities for micropalaeontological research. We have been using the image analysis program NIH-Image extensively over the last few years in coccolith research and have developed macro program solutions to a range of problems. Our current work is focussed on development and application of these macros in the CODENET project. A selection of these macros are made available here.

        What are coccoliths? - Advice to workers who use NIH-Image but have no interest in coccoliths
        • Coccoliths are cute little calcareous plates produced by a group of planktonic algae, for more information see the INA Web site.
        • The mosaic-archiving macros might be of use in any Image application were the individual objects being studied are small (<100 pixels).
        • The biometrics macros are specialised but might give ideas or even a macro framework for other biometric applications.
        • Advanced biometrics is not really possible on coccoliths, essentially these macros are used for measuring a few simple size and shape parameters. Numerous studies have shown that these types of measurements are valuable, but this is not comparable to the types of advanced morphometric analysis used on many other organisms.


        What is NIH-Image? - Advice to coccolith workers who have never used the program
        NIH-Image is a widely used image analysis program developed by Wayne Rasband of the National Institutes of Heath (USA) and available as public domain software. The program was developed for Apple Macintosh computers and our macros only work reliably on Apple MacIntoshes. A version for Windows - Scion-Image- is at an advanced stage of development... but the macro language does not work predictably so we cannot recommend it for our type of use yet (late 1999). It is freely available from Scion Corporation.
        Image for OSX - unfortunately NIH-Image does not work in OSX, even in OS9 emulation, however, Steve Barrett of Liverpool University, England, is developing a version of NIH-Image "Image-SXM" for OSX. At present (July 2003) it is at beta testing stage with some bugs, but basically it works. To obtain a copy contact Steve by email - S.D.Barrett@liv.ac.uk

        NIH-Image works with pixel based images, and has many similar features to image manipulation programs such as Photoshop. In addition:
        • It has the ability to directly drive a range of framegrabbers so that it can be used for image acquisition.
        • It contains a suite of morphometric routines allowing measurement of features on images, with results being saved typically to a spreadsheet program such as Excel for analysis.
        • It incorporates a Pascal -like macro programming language allowing automation of tasks and development of custom applications (see also notes on the NIH Image Macro Language).

        How to use our macro sets
         We have developed our applications as sets of macros. Within each macro set individual macros carry out specific tasks, these appear as individual commands within a menu of NIH-Image, and can be called via the keyboard. Different macros within a macro set complement each other to provide a customised solution to a particular research task.
        To use the macros provide here it is necessary to:
        1. Download and install the program NIH-Image or Scion Image. Homesite for NIH-Image (Mac versions only). Homesite for Scion Image (Windows and Mac versions).
        2. Download the relevant macro sets (N.B. Download as text and avoid adding extra carriage returns).
        3. Run the program and load the macro set.
        4. Experiment - we have included some demonstration images on this site to aid experimentation.
        See also -Notes on downloading the macros and microscope calibration.

        OUR MACRO-SET APPLICATIONS

        CoccoBiom
        This is the main macro set we have developed, it is designed for working with a light microscope system consisting of microscope and CCD camera to produce images, together with a framegrabber in a computer to capture the images. The macros within the set provide a range of functions for image capture and measurement. See separate page for details.

        SEM techniques
        Routines are being developed, in association with colleagues at ETH-Zurich for use with nannofossil images captured from digital Scanning Electron Microscopes. These include:
        1. Macros for counting frequency of specimens in a set of images captured by an automated electron microscope. The images are automatically opened in sequence and different species can be enumerated separately.
        2. Macros for automatically printing out reference sets of images at convenient magnification for archiving.
        Queries on these to Markus Geisen.
        Volume of rotation calculation
        Coccolith volume/mass estimates are valuable for analysis of sediment trap data - to calculate coccolith carbonate fluxes. Since coccoliths have a high degree of rotational symmetry their volumes can be calculated from cross-sections. We have used NIH-Image to perform such calculations. The macros take as their input cross sections and axial ratios and calculate volumes and masses (assuming the coccolith is formed of calcite. This application is documented in Young & Ziveri (in press). The macros used are provided here.

        Rock colour measurement
        Rock colour is increasingly being used as a very readily measured proxy for carbonate content. Most such work is based on scanning of slabbed cores, this is for instance becoming a routine tool in analysis of ODP cores. From outcrop sections similar data sets can be generated by:
        (1) Collecting closely spaced sample sets.
        (2) Grinding a smooth surface on the samples.
        (3) Oven drying them.
        (4) Placing the samples on a flat bed scanner and scanning them.
        (5) Measuring the rock colour using an image analysis program.
        We have applied this technique to rhythm analysis in pelagic sediments showing Milankovitch cyclicity using simple greyscale variation (Gale et al submitted). The technique is documented in detail in Young et al. (submitted).The macro set used for the greyscale measurement is given here, this is a very simple macro set, but perhaps a useful introduction to the macro language.

        Return to: top,
        CODENET WWW Site Index Page
        References

        This page is based on work primarily of Jeremy Young and Markus Geisen
        Comments and queries are always welcome