作者: Sezer Karaoglu , Sajib Kumar Saha , Basura Fernando
DOI:
关键词: Pattern recognition 、 Translation (geometry) 、 Computer vision 、 Scale (ratio) 、 Point (geometry) 、 Object detection 、 Affine transformation 、 Rotation (mathematics) 、 Random forest 、 Interpolation 、 Artificial intelligence 、 Object-class detection 、 Computer science
摘要: This paper presents a novel multi scale gradient and corner point based shape descriptors. The descriptor is combined with generic Fourier descriptors to extract contour region information. Shape information object class detection classification technique random forest classifier has been optimized. Proposed integrated in this robust rotation, scale, translation, affine deformations, noisy contours shapes. new interpolated exploited for fast higher accuracy.