QUICK GUIDE TO FUNDAMENTAL PROCEDURES IN IMAGE PRO PLUS

CALIBRATION OF IMAGES

INTERACTIVE MEASUREMENT OF METRICS FROM A CALIBRATED IMAGE

CREATION OF AN AREA OF INTEREST (AOI)

SEGMENTATION OF PIXELS WITHIN AN AOI


CALIBRATION OF IMAGES

In order for your measurements to be meaningful, you need to make sure that your images are calibrated.  This usually involves capturing an image of a stage micrometer scale with the same optical set up as your data images.  Ideally you capture calibration images each time you capture data images, since parameters can be different between data collection periods.  Once you have calibrated a group of data images, you can analyze them by reloading your saved calibration file associated with those images.  Obviously accurate record keeping is a must!

I.  Load Stage Micrometer Scale Image

Select:
 
Measure
    Calibration {spatial}
        Spatial (set name, Unit = microns)
        Image [Adjust scale to span xx microns]
            {Make sure you stretch edge to comparable edge!}

    Calibration {gray level}
        Intensity (set name, Unit = OD)
        Std Optical Density
            Options (set Black level = 0, Incident level = 255)

    Calibration {save}
        Save active

INTERACTIVE MEASUREMENT OF METRICS FROM A CALIBRATED IMAGE

I. Load data image
II. Load calibration file for this image

Measure
    Measurement
        Select line feature icon (or some other icon)
        Move/Stretch icon to features

CREATION OF AN AREA OF INTEREST (AOI)

An area of interest (AOI) delimits the population of pixels in your image which you want to segment on the basis of gray levels.  In my lab, AOI usually equates to organs, or spatial regions within organs - eg. For stems - epidermis, cortex, vascular bundles, medullary rays, pith.  For leaves - epidermis, hypodermis, mesophyll, vascular bundles.  For roots - epidermis, cortex, endodermis, pericycle, xylem, phloem.  AOI's are usually humanly defined categories that are of interest to a research study - in my experience, they are not entities that naturally emerge from computer analysis.  So this is the most interactive aspect of most quantitative digital image analysis research projects.

I. Load data image
II. Load calibration file for this image

New AOI

    Irregular AOI
        Trace<>Wand
            Wand (set Thresh=3, Smooth=0, Speed=5, Noise=5)

            LMB =  first point & to free hand draw
            LMB (twice) = auto trace
            RMB (hold, drag away) = to correct auto trace
            RMB = convert to AOI

            {The above is tricky!  To do it right - i.e. make the machine do your bidding, as opposed to the machine driving you, takes some practice!}

    Edit AOI
        AOI
            Add {Rename Polygon to appropriate name}
            Save AOI

SEGMENTATION OF PIXELS WITHIN AN AOI

I. Load data image
II. Load calibration file for this image
III.  Load the AOI file

    Edit AOI, select appropriate AOI file
    Within the AOI window, select AOI for analysis

IV.  Click Measure –> Count/size

    Measure Menu –> Select Measurements
        Set all the measurements that you want to make on the thesholded objects

    Set Options in count/size dialog box:

    outline = filled, Label style, Label color, 8-connect, smoothing = 0,
    Clean Borders = none <<<< NB.:Important!

V.  Theshold the pixels to be analyzed

    Click "Select Colors"
        Use pointer to pick an area with the right color

    OR

    Select Histogram Based tab
    [It's usually easier to use HVI rather than RGB]
        Set the upper and lower limits for each HVI channel, thesholded pixels will be displayed on image

    NB.: If you want the entire AOI measured, set all channels to 0 -255

VI.  Click Count

    Theholded pixels and objects in AOI will be outlined and labeled on image

    DDE to export measurements to excel file

VII. To view measurements