Open AccessData‐driven nonrigid object feature analysis: Subspace application of incidence structure
Author(s) -
Wells Nicholas,
See Chung W.
Publication year - 2020
Publication title -
engineering reports
Language(s) - English
Resource type - Journals
ISSN - 2577-8196
DOI - 10.1002/eng2.12141
Subject(s) - artificial intelligence , pattern recognition (psychology) , cognitive neuroscience of visual object recognition , hough transform , computer science , affine transformation , clutter , feature extraction , feature (linguistics) , object (grammar) , subspace topology , 3d single object recognition , computer vision , linear discriminant analysis , mathematics , image (mathematics) , radar , geometry , linguistics , philosophy , telecommunications
In object recognition, feature extraction algorithms are designed to capture the discriminate statistics of objects. Due to pose, deformation and background clutter, the recognition of objects becomes nontrivial, particularly nonrigid samples. Through incidence and geometric structure, this article reports on the data‐driven identification of critical features located on object exemplar profiles. The investigation is demonstrated using the features of a cat's head and the application of the Hough transform to extract planar geometric features. A data‐driven recognition routine is described that accumulates prior knowledge for evaluating the error contribution of critical features impacting recognition confidence. In measuring affine facial features, feature parallelism is tracked to determine rotations and elevations of a cat's head. A preliminary recognition error of 8.2% and 17.8% is determined for a front training profile. An analysis proceeds to determine contributions to this error due the identified critical features.