Detection and identification of elliptical structure arrangements in images : theory and algorithms

摘要: This thesis deals with different aspects concerning the detection, fitting, and identification of elliptical features in digital images. We put geometric feature detection a contrario statistical framework order to obtain combined parameter-free line segment, circular/elliptical arc detector, which controls number false detections. To improve accuracy detected features, especially cases occluded circles/ellipses, simple closed-form technique for conic fitting is introduced, merges efficiently algebraic distance gradient orientation. Identifying configuration coplanar circles images through discriminant signature usually requires Euclidean reconstruction plane containing circles. propose an efficient computation method that bypasses reconstruction; it relies exclusively on invariant properties projective plane, being thus itself under perspective.