Curvilinear object segmentation with geometric priors
AssigneePalo Alto Research Center Incorporated
InventorsRaja Bala, Venkateswararao Cherukuri, Vijay Kumar B G
Life of this patent
Estimating the term…
Abstract
A method for curvilinear object segmentation includes receiving at least one input image comprising curvilinear features. The at least one input image is mapped to segmentation maps of the curvilinear features using a deep network having a representation module and a task module. The mapping includes transforming the input image in the representation module using learnable filters configured to balance recognition of curvilinear geometry with reduction of training error. The segmentation maps are produced using the transformed input image in the task module.
Claims22
A method for curvilinear object segmentation, comprising:
- receiving at least one input image comprising curvilinear features;
- and
- mapping, using a processor, the at least one image to segmentation maps of the curvilinear features using a deep network having a representation module and a task module, the mapping comprising:
- transforming the input image in the representation module using learnable filters configured to balance recognition of curvilinear geometry with reduction of training error;
- and
- producing the segmentation maps using the transformed input image in the task module.
A system for curvilinear object segmentation, comprising:
- a processor;
- and
- a memory storing computer program instructions which when executed by the processor cause the processor to perform operations comprising:
- receiving at least one input image comprising curvilinear features;
- and
- mapping, using a processor, the at least one image to produce segmentation maps of the curvilinear features using a deep network having a representation module and a task module, the mapping comprising:
- transforming the input image in the representation module using learnable filters configured to balance recognition of curvilinear geometry with reduction of training error;
- and
- producing the segmentation maps using the transformed input image in the task module.
A non-transitory computer readable medium storing computer program instructions for curvilinear object segmentation, the computer program instructions when executed by a processor cause the processor to perform operations comprising:
- receiving at least one input image comprising curvilinear features;
- and
- mapping, using a processor, the at least one image to produce segmentation maps of the curvilinear features using a deep network having a representation module and a task module, the mapping comprising:
- transforming the input image in the representation module using learnable filters configured to balance recognition of curvilinear geometry with reduction of training error;
- and
- producing the segmentation maps using the transformed input image in the task module.
Dependencies are inferred from the claim text.