System and method for dynamic image recognition

A method of creating a knowledge base for use in image recognition for images on a set of derived spaces, comprising:

• imaging at least one blank object;
• creating a presence/absence knowledge base by:
• generating a set of presence/absence features for each derived space from at least one blank object;
• scoring each presence/absence feature for each derived space from a first blank object;
• and
• scoring each presence/absence feature for each derived space from every other blank object and every assembled object imaged;
• creating a polarity knowledge base by:
• generating a set of polarity features for each derived space from a first assembled object;
• rotating the each feature image to generate a set of wrong polarity features;
• and
• scoring each polarity feature for each derived space for every assembled object imaged.

A method for dynamic image recognition for making determinations about regions of interest on an object using an image recognition computer program contained in computer-readable form on a tangible storage medium, comprising the steps of executing the image recognition computer program to:

• acquiring a raw gray level image of the object comprising a set of raw pixels, each raw pixel having a gray level value;
• derive a number of distinct spaces from the raw pixel image;
• cluster the raw pixel images for each distinct space based on a set of clustering rules;
• and
• generate a set of features for each region of interest in each distinct space, each feature having an arbitrary geometry.

A method for dynamic image recognition, comprising:

• collecting raw image data from at least one imaged object;
• segmenting out a region of interest;
• performing at least one spatial image transform to generate a set of derived spaces for the region of interest;
• and
• clustering the region of interest to generate a set of features for each derived space based on a set of clustering rules to generate arbitrary features on the region of interest for each derived space.

A method for performing basic training in a dynamic image recognition system, comprising:

• collecting raw image data from at least one blank object;
• segmenting out a region of interest on the at least one blank object;
• performing at least one spatial image transform to generate a set of derived spaces for the region of interest;
• generating a set of presence/absence features for the at least one blank object, comprising;
• spatially clustering similar presence/absence pixel codes in each derived space to create a set of presence/absence features;
• and
• representing each significant presence/absence feature as one or more presence/absence boxes;
• scoring each presence/absence box on each blank object, thereby building a blank object presence/absence score set;
• collecting raw image data from at least one assembled object;
• segmenting out a region of interest on the at least one assembled object;
• performing at least one spatial image transform to generate a set of derived spaces for the region of interest;
• scoring each present/absence box on each assembled object, thereby building an assembled board presence/absence score set;
• building a presence/absence knowledge base using the blank object presence/absence score set and the assembled object presence absence score set;
• pruning each presence/absence box;
• generating a set of polarity features from a first assembled object, comprising;
• spatially clustering similar polarity-pixel codes in each derived space to create at least one polarity feature;
• and
• representing each significant polarity feature as one or more polarity boxes;
• scoring each polarity box on each assembled object, thereby building an assembled object score set;
• building a polarity knowledge base using the assembled object score set;
• and
• pruning each polarity box.

An image recognition computer software program for dynamic image recognition for making determinations about regions of interest on an object, the image recognition computer program stored in a computer-readable form on a tangible storage medium and executable by a processor to:

• load a knowledge base containing a set of features and past decision confidences for each region of interest for the object to be imaged;
• acquire a raw gray level image of the object comprising a set of raw pixels, each raw pixel having a gray level value;
• derive a number of distinct spaces from the raw pixel image;
• score each feature for each region of interest on each of the derived spaces;
• classify each feature for both presence/absence and polarity;
• calculate decision confidence for each feature;
• and
• calculate a relative decision confidence for each feature to provide a metric that compares the decision confidence for each feature to past decision confidences for the feature contained in the knowledge base;
• and
• if the relative decision confidence is outside of a predetermined window, classify the feature as a defect and perform an extended relative confidence analysis to subclassify the defect.

A knowledge base creation computer software program stored on in computer-readable format on a tangible storage medium, the program for use in image recognition for images on a set of derived spaces using at least one bland object and one assembled object, the program executable by a processor to:

• create a presence/absence knowledge base by:
• generating a set of presence/absence features for each derived space from at least one blank object;
• scoring each presence/absence feature for each derived space from a first blank object;
• and
• scoring each presence/absence feature for each derived space from every other blank object and each assembled object imaged;
• create a polarity knowledge base by:
• generating a set of polarity features for each derived space from a first assembled object;
• rotating the each feature image to generate a set of wrong polarity features;
• and
• scoring each polarity feature for each derived space for every assembled object imaged.

A system for dynamic image recognition for making determinations about regions of interest on an object, comprising:

• a processor;
• and
• an image recognition computer program stored in computer-readable form on a tangible storage medium, the image recognition computer program executable to:
• acquire a raw gray level image of the object comprising a set of raw pixels, each raw pixel having a gray level value;
• derive a number of distinct spaces from the raw pixel image;
• cluster the raw pixel images for each distinct space based on a set of clustering rules;
• and
• generate a set of features for each region of interest in each distinct space, each feature having an arbitrary geometry.

A system for dynamic image recognition, comprising:

• a processor;
• and
• an image recognition computer software program stored in computer-readable form on a storage medium and executable to:
• collect raw image data from at least one imaged object;
• segment out a region of interest;
• perform at least one spatial image transform to generate a set of derived spaces for the region of interest;
• and
• cluster the region of interest to generate a set of features for each derived space based on a set of clustering rules to generate arbitrary features on the region of interest for each derived space.

A system for performing basic training in a dynamic image recognition system, comprising:

• a processor;
• and
• a training computer software program stored in computer-readable form on a storage medium and executable to:
• collect raw image data from at least one blank object;
• segment out a region of interest on the at least one blank object;
• perform at least one spatial image transform to generate a set of derived spaces for the region of interest;
• generate a set of presence/absence features for the at least one blank object, comprising;
• spatially clustering similar presence/absence pixel codes in each derived space to create a set of presence/absence features;
• and
• representing each significant presence/absence feature as one or more presence/absence boxes;
• score each presence/absence box on each blank object, thereby building a blank object presence/absence score set;
• collect raw image data from at least one assembled object;
• segment out a region of interest on the at least one assembled object;
• perform at least one spatial image transform to generate a set of derived spaces for the region of interest;
• score each present/absence box on each assembled object, thereby building an assembled board presence/absence score set;
• build a presence/absence knowledge base using the blank object presence/absence score set and the assembled object presence absence score set;
• prune each presence/absence box;
• generate a set of polarity features from a first assembled object, comprising;
• spatially clustering similar polarity pixel codes in each derived space to create at least one polarity feature;
• and
• representing each significant polarity feature as one or more polarity boxes;
• score each polarity box on each assembled object, thereby building an assembled object score set;
• build a polarity knowledge base using the assembled object score set;
• and
• prune each polarity box.