This extensive tutorial is included as a PDF (Adobe Acrobat) file on the Fovea Pro and Image Processing Tool Kit CD-ROMs. It is intended to provide a hands-on guide to learning how to perform image processing and measurement procedures. The user may either consult it as a reference when trying to learn a specific task, or work systematically through it, in which case it functions as a college-level course in the subject (and has been used for that purpose many times).

Note: The online version of this document has been divided into 22 parts, below. The PDF version runs about 150 printed pages.

Contents

• Part 1
  1. The program environment
  2. Correction of image defects
  2.A. Color images
  2.A.1 Is color correction required?
  2.A.2 Color filtering and channel separation to improve contrast
  2.A.3 Merging color channels

• Part 2
  2.A.4 Principal Components Analysis

• Part 3
  2.B. Noisy images
  2.B.1 Random speckle noise
  2.B.2 Shot noise and scan line noise removal

• Part 4
  2.B.3 Periodic noise removal

• Part 5
  2.C. Nonuniform image illumination
  2.C.1 Is a separate background image available?
  2.C.2 Is the background visible throughout the image?
  2.C.3 Correcting varying contrast across an image
  2.C.4 Are the features small in one dimension?

• Part 6
  2.D. Expanding image contrast
  2.D.1 Linear expansion for grayscale and color images
  2.D.2 Non-linear adjustments (gamma, equalization)
  2.E. Distorted or foreshortened images
  2.E.1 Making pixels square
  2.E.2 Perspective distortion (non perpendicular viewpoint)

• Part 7
  2.F. Focus problems
  2.F.1 Shallow depth of field
  2.F.2 Deconvolution of blurred focus
  2.G. Tiling large images
  2.G.1 Shift and align multiple fields of view
  3. Enhancement of image detail
  3.A. Poor local contrast and faint boundaries or detail
  3.A.1 Local equalization

• Part 8
  3.A.2 Sharpening (high pass filters)
  3.A.3 Unsharp mask and difference of Gaussians
  3.A.4 Color images should be processed in HSI space
  3.A.5 Feature selection based on size

• Part 9
  3.A.6 Pseudo-color, pseudo-3D, and other display tools
  3.B. Are feature edges important?
  3.B.1 Edge enhancement with derivative operators
  3.B.2 Increasing edge sharpness and region uniformity

• Part 10
  3.C. Converting texture and directionality to grayscale or color differences

• Part 11
  3.D. Fourier-space processing
  3.D.1 Isolating periodic structures or signals
  3.D.2 Location of specific features
  3.E. Other uses of correlation
  3.E.1 Alignment
  3.E.2 Measurement of fractal dimension

• Part 12
  3.F. Detecting image differences
  3.F.1 Alignment
  3.F.2 Subtraction and ratioing
  3.G. Summary of image processing procedures
  4. Thresholding of image features
  4.A. Thresholding using the histogram
  4.A.1 Manual settings

• Part 13
  4.A.2 Automatic methods
  4.A.3 Selecting a color range

• Part 14
  4.A.4 Contour lines (iso-brightness lines)
  4.B. Marking features manually
  4.B.1 Region growing
  4.B.1 Manual lines, points, etc.

• Part 15
  5. Binary image processing
  5.A. Removing extraneous lines, points, or other features
  5.A.1 Erosion/dilation with appropriate coefficients to remove lines or points
  5.A.2 EDM-based morphology to remove small features or protrusions, and fill in gaps
  5.B. Separating features that touch

• Part 16
  5.C. Combining multiple images of the same area to select features
  5.C.1 Boolean logic to apply multiple criteria

• Part 17
  5.C.2 Using markers to select objects
  5.C.3 Region outlines as selection criteria

• Part 18
  5.D. Feature skeletons provide important shape characterization
  5.D.1 Grain boundary, cell wall, and fiber images
  5.D.2 Measuring skeleton length, number of ends, number of branches for features
  5.E. Using the Euclidean distance map for feature measurements
  5.E.1 Distances from boundaries or objects

• Part 19
  6. Measurements
  6.A. Calibration
  6.A.1 Calibrating image dimensions
  6.A.2 Calibrating density/grayscale values
  6.B. Measurement of principal stereological parameters:
  6.B.1 Volume fraction
  6.B.2 Surface area
  6.B.3 Line length
  6.B.4 Intercept lengths
  6.B.5 Number per unit volume
  6.B.6 The Disector
  6.B.7 Point-sampled intercepts
  6.B.8 Other stereological methods

• Part 20
  6.C. Intensity and Color
  6.D. Counting features

• Part 21
  6.E. Measuring features
  6.F. Data analysis and feature recognition

• Part 22
  7. Automation
  7.A. Photoshop Actions and FoveaPro Logging
  7.B. Batch Processing
  7.C. Scripting

The tutorial and FoveaPro are natural complements to John C. Russ, The Image Processing Handbook (5th edition, CRC Press, 2006, ISBN 0-8493-7254-2).

Other recommended texts include:

• John C. Russ, Forensic Uses of Digital Imaging (CRC Press, 2001 ISBN 0-8493-0903-4)
• John C. Russ, Image Analysis of Food Microstructure (CRC Press, 2004 ISBN 0-8493-2241-3)
• John C. Russ & Robert T. Dehoff, Practical Stereology (2nd edition, Plenum Press, 2001, ISBN 0-306-46476-4)
  The stereological measurement methods are particularly important for correctly inferring 3D shape and structure from 2D images.

You can order these widely used textbooks and handbooks from Amazon.com. All of the images shown in the tutorial are included on the Fovea Pro and Image Processing Tool Kit CDs.

If you want more information on these and other related textbooks, our documentation section has an extensive bibliography.