Chapter 7: Geometric Model Finder

This chapter explains how to use the Geometric Model Finder module to locate occurrences of models in your target.

• Geometric Model Finder module
• Steps to performing a model search
• Basic concepts for the MIL Geometric Model Finder module
• Types of Model Finder contexts
• • General geometric contexts
  • Controlled geometric contexts
  • Circle shape contexts
• Defining and adding models to your Model Finder context
• • Image-type models
  • • Extracting edges
  • Synthetic models
  • • Predefined-shape models
    • CAD-file models
    • Model size and units
    • Synthetic models and calibration
  • Models defined from result buffers
  • • Models defined from an Edge Finder result buffer
    • Models defined from a Model Finder result buffer
  • Models defined from two other models
  • Preprocessing
  • Model origin
  • Model indices and labels
  • Drawing and inquiring the model's active edges
• Guidelines for choosing models
• • Make sure your images have enough contrast
  • Avoid poor geometric models
  • Be aware of ambiguous models
  • Nearly ambiguous models
• Masking your model
• Search targets
• • Finding models in an image
  • Finding models in an Edge Finder result buffer
• Determining what is a match
• • Score and target score
  • Model and target coverage
  • Fit error
  • Interpreting results
• Position, angle, and scale
• • Enabling calculations specific to searching within a range
  • Search position and position range
  • Angle and angular range
  • Scale and scale range
• Customizing search settings
• • Acceptance levels
  • Certainty levels
  • Expected number of occurrences
  • Reference axis
  • • Reference axis origin
    • Reference axis angle
    • Forward and reverse transformation coefficients
• Advanced search settings
• • Polarity
  • Radii deviation tolerance
  • Separation
  • Shared edges
  • Fit error weighting factor
• Global context settings
• • Setting the search speed
  • Accuracy
  • Timing out your search
  • First and last levels
• Speeding up the search
• • Adjust the search speed of the algorithm
  • Disable calculations specific to range search strategies
  • Define several models with different expected angles
  • Limit the position range
  • Limit the search scale
  • Specify the exact number of expected occurrences
  • Clean up your model and your target
  • Changing the search levels
  • Using an appropriate context type for your search
• Retrieving and analyzing results
• • Possible results
  • Drawing results
• Calibration
• • Requirements
  • Models and their calibration context
  • Setting the aspect ratio control
  • Using transformation coefficients with calibrated images
• Geometric Model Finder example