摘要:
An alignment template goodness qualification method receives a pattern image and a pattern based alignment template and performs template goodness measurement using the pattern image and the pattern based alignment template to generate template goodness result output. A template qualification is performed using the template goodness result to generate template qualification result output. If the template qualification result is acceptable, the pattern based alignment template is outputted as the qualified pattern based alignment template. Otherwise, an alternative template selection is performed using the pattern image, the pattern based alignment template and the template goodness result to generate alternative pattern based alignment template output. The template goodness measurements include signal content measurement, spatial discrimination measurement and pattern ambiguity measurement.
摘要:
A partition pattern template generation method for alignment receives a learning image and performs partition template generation using the learning image to generate a plurality of partition template result output. A partition template acceptance test is performed using the plurality of partition template results to generate partition templates or failure result. A partition template search method for alignment receives an alignment image and partition templates and performs a plurality of template search steps to generate a plurality of matching scores output. A partition integration method is performed using the plurality of matching scores to generate a partition template search result. A partition integration error self checking method receives a preliminary template search result position and a plurality of the matching scores. A matching score profile comparison is performed using the plurality of the matching scores and the expected matching score profile to generate the template search result.
摘要:
A fast high precision matching method receives an input image and a template. An initial search method uses the input image and the template to create an initial search result output. A high precision match uses the initial search result, the input image, and the template to create a high precision match result output. The high precision match method estimates high precision parameters by image interpolation and interpolation parameter optimization. The high precision match method also performs robust matching by limiting pixel contribution or pixel weighting. An invariant high precision match method estimates subpixel position and subsampling scale and rotation parameters by image interpolation and interpolation parameter optimization on the log-converted radial-angular transformation domain. This invention provides a fast method for high precision matching with the equivalent subpixel and subsampling interpolation in the image or template domain without actual performing the subpixel interpolation and/or subsampling. It achieves the high precision through sampling parameter optimization. Therefore, very fine sampling precision can be accomplished without the difficulty of high resolution image/template storage and expensive computation for actual matching at high resolution. This invention is generalized to include the high precision scale and rotation invariant matching through parameter optimization on log-converted radial-angular coordinate. This invention can be easily generalized to three-dimensional or higher dimensional invariant high precision pattern search and can achieve even greater speed advantage comparing to the prior art methods. Therefore, it can be used in applications such as 3D medical imaging, dynamic medical imaging, confocal microscopy, live cell assays in drug discovery, or ultrasound imaging.
摘要:
A directed pattern enhancement method receives a learning image and pattern enhancement directive. Pattern enhancement learning is performed using the learning image and the pattern enhancement directive to generate pattern enhancement recipe. An application image is received and a pattern enhancement application is performed using the application image and the pattern enhancement recipe to generate pattern enhanced image. A recognition thresholding is performed using the pattern enhanced image to generate recognition result. The pattern enhancement directive consists of background directive, patterns to enhance directive, and patterns to suppress directive. A partitioned modeling method receives an image region and performs feature extraction on the image region to generate characterization feature. A hierarchical partitioning is performed using the characterization feature to generate hierarchical partitions. A model generation is performed using the hierarchical partitions to generate partition model. The partitioned modeling further performs a partitioned matching step that matches an input point to the partition model to generate a matching score output. A partition model update method receives a partition model and input data for model update. A partition model update is performed using the partition model and the data to generate an updated partition model.
摘要:
A feature regulation application method for hierarchical decision learning systems receives a feature regulation training data. A feature regulation method uses the feature regulation training data and invokes a plurality of the hierarchical decision learning to create feature subset information output. The feature regulation application method also receives a learning data. A feature sampling method uses the feature subset information and the learning data to create a feature subset learning data output. A hierarchical decision learning method uses the feature subset learning data to create a hierarchical decision system output. The feature regulation method also outputs feature ranking information. A feature regulated hierarchical decision learning method uses the feature subset learning data and the feature ranking information to create a hierarchical decision system output. This invention performs feature selection using a feature regulation method designed specifically for hierarchical decision learning systems such as decision tree classifiers. It provide a computationally feasible method for feature selection that considers the hierarchical decision learning systems used for decision making. It evaluates the stability of features subject to context switching and the reliability of the tree nodes by information integration. It provides the ranking of the features that can be incorporated in the creation of the hierarchical decision learning systems.
摘要:
An object based boundary refinement method for object segmentation in digital images receives an image and a single initial object region of interest and performs refinement zone definition using the initial object regions of interest to generate refinement zones output. A directional edge enhancement is performed using the input image and the refinement zones to generate directional enhanced region of interest output. A radial detection is performed using the input image the refinement zones and the directional enhanced region of interest to generate radial detection mask output. In addition, a final shaping is performed using the radial detection mask having single object region output. A directional edge enhancement method determining pixel specific edge contrast enhancement direction according to the object structure direction near the pixel consists receives an image and refinement zones and performs 1D horizontal distance transform and 1D vertical distance transform using the refinement zones to generate horizontal distance map and vertical distance map outputs. A neighboring direction determination is performed using the horizontal distance map and the vertical distance map to generate neighboring image output. In addition, a directional edge contrast calculation using the neighboring image and input image having directional enhanced region of interest output.
摘要:
An intelligent spatial reasoning method receives a plurality of object sets. A spatial mapping feature learning method uses the plurality of object sets to create at least one salient spatial mapping feature output. It performs spatial reasoning rule learning using the at least one spatial mapping feature to create at least one spatial reasoning rule output. The spatial mapping feature learning method performs a spatial mapping feature set generation step followed by a feature learning step. The spatial mapping feature set is generated by repeated application of spatial correlation between two object sets. The feature learning method consists of a feature selection step and a feature transformation step and the spatial reasoning rule learning method uses the supervised learning method.The spatial reasoning approach of this invention automatically characterizes spatial relations of multiple sets of objects by comprehensive collections of spatial mapping features. Some of the features have clearly understandable physical, structural, or geometrical meanings. Others are statistical characterizations, which may not have clear physical, structural or geometrical meanings when considered individually. A combination of these features, however, could characterize subtle physical, structural or geometrical conditions under practical situations. One key advantage of this invention is the ability to characterize subtle differences numerically using a comprehensive feature set.
摘要:
A structure-guided transformation transforms a region of an image into a region in the structure-transformed image according to the desired structure. The invention achieves efficient and accurate structure-guided processing such as filtering, detection and comparison in the transformed domain and thereby facilitates use of simple operations to enhance or detect straight lines or edges. Structure information is used to enhance and detect image features of interest even when the shape of the image structure is not regular. Both global and local structures of objects can be inspected. Global structure inspection detects gross errors in image structure; therefore side effects caused by mismatched structure-guided processing are avoided. Subtle defects along the edge of a structure can be detected by local structure inspection. Structure information guidance provides an edge detection inspection system that tolerates significant noise and contrast variations.
摘要:
Mark detection and position determination are improved by use of directional elongated filters, symmetry, gray scale image processing, structural constraints, and learning. Directional elongated filters are used to pre-process images of registration marks to create masks and enhanced images. Working sequentially, portions of the mark are detected and classified. The input gray scale image of the mark is processed using its structural constraints in conjunction with a mask for the detected mark. A cost function estimation determines mark position and orientation with sub-pixel accuracy. Learning is used to improve specific application performance.
摘要:
A high speed image processing apparatus is created through the use of cascaded elongated filters. The processing speed of the filters is kernel size insensitive, enabling use of general purpose computing facilities to process high resolution, monochrome, and multi-spectrum images. Elongated filters described include both linear and non-linear filters. Very large kernel and multi-dimensional image processing is accomplished with reduced complexity and portable programming instructions.