Results 1 -
4 of
4
Spatial Information Retrieval from Images using Ontologies and Semantic Maps
"... Abstract. Cameras provide integrated GPS technology which makes them a powerful sensor for geographical context related images. They allow wire-less connection to computers and the images can be automatically trans-ferred to a PC or can be integrated into a GIS system. In this paper we pro-pose an a ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
(Show Context)
Abstract. Cameras provide integrated GPS technology which makes them a powerful sensor for geographical context related images. They allow wire-less connection to computers and the images can be automatically trans-ferred to a PC or can be integrated into a GIS system. In this paper we pro-pose an approach for spatial information retrieval from images using the concept of ontologies and semantic maps. The term of ontology is used in our case to describe spatial domain knowledge to enhance the search capa-bility and image annotation. The objects are represented by their location in semantic maps. We describe a developed prototype system with a database design for ontologies and semantic maps. We demonstrate the automatic image annotation and the visualization of the spatial queries. The system is oriented to the area of culture and tourism and provides a user friendly in-terface.
unknown title
"... In this paper, we propose a method, called the nearest feature midpoint (NFM), for pattern classifica-tion. Any pair of feature points of the same class is generalized by the feature midpoint (FM) between them. Hence the representational capacity of available prototypes can be expanded. The classifi ..."
Abstract
- Add to MetaCart
(Show Context)
In this paper, we propose a method, called the nearest feature midpoint (NFM), for pattern classifica-tion. Any pair of feature points of the same class is generalized by the feature midpoint (FM) between them. Hence the representational capacity of available prototypes can be expanded. The classification is determined by the nearest distance from the query feature point to each FM. This paper compares the NFM classifier against the nearest feature line (NFL) classifier, which has reported successes in various applications. In the NFL, any pair of feature points of the same class is generalized by the feature line (FL) passing through them, and the classification is evaluated on the nearest distance from the query feature point to each FL. The NFM can be considered to be the refinement of the NFL. A theoretical proof is provided in this paper to show that for the n-dimensional Gaussian distribution, the classification based on the NFM distance metric will achieve the least error probability as com-pared to those based on any other points on the feature lines. Furthermore, a theoretical investigation is provided that under certain assumption the NFL is approximately equivalent to the NFM when the dimension of the feature space is high. The experimental evaluations on both simulated and real-life benchmark data concur with all the theoretical investigations, as well as indicate that the NFM is ef-fective for the classification of the data with a Gaussian distribution or with a distribution that can be reasonably approximated by a Gaussian.
SHREC’15 Track: Non-rigid 3D Shape Retrieval†
"... Non-rigid 3D shape retrieval has become a research hotpot in communities of computer graphics, computer vi-sion, pattern recognition, etc. In this paper, we present the results of the SHREC’15 Track: Non-rigid 3D Shape Retrieval. The aim of this track is to provide a fair and effective platform to e ..."
Abstract
- Add to MetaCart
(Show Context)
Non-rigid 3D shape retrieval has become a research hotpot in communities of computer graphics, computer vi-sion, pattern recognition, etc. In this paper, we present the results of the SHREC’15 Track: Non-rigid 3D Shape Retrieval. The aim of this track is to provide a fair and effective platform to evaluate and compare the perfor-mance of current non-rigid 3D shape retrieval methods developed by different research groups around the world. The database utilized in this track consists of 1200 3D watertight triangle meshes which are equally classified into 50 categories. All models in the same category are generated from an original 3D mesh by implementing vari-ous pose transformations. The retrieval performance of a method is evaluated using 6 commonly-used measures (i.e., PR-plot, NN, FT, ST, E-measure and DCG.). Totally, there are 37 submissions and 11 groups taking part in this track. Evaluation results and comparison analyses described in this paper not only show the bright future in researches of non-rigid 3D shape retrieval but also point out several promising research directions in this topic.
unknown title
"... In this paper, we propose a method, called the nearest feature midpoint (NFM), for pattern classifica-tion. Any pair of feature points of the same class is generalized by the feature midpoint (FM) between them. Hence the representational capacity of available prototypes can be expanded. The classifi ..."
Abstract
- Add to MetaCart
(Show Context)
In this paper, we propose a method, called the nearest feature midpoint (NFM), for pattern classifica-tion. Any pair of feature points of the same class is generalized by the feature midpoint (FM) between them. Hence the representational capacity of available prototypes can be expanded. The classification is determined by the nearest distance from the query feature point to each FM. This paper compares the NFM classifier against the nearest feature line (NFL) classifier, which has reported successes in various applications. In the NFL, any pair of feature points of the same class is generalized by the feature line (FL) passing through them, and the classification is evaluated on the nearest distance from the query feature point to each FL. The NFM can be considered to be the refinement of the NFL. A theoretical proof is provided in this paper to show that for the n-dimensional Gaussian distribution, the classification based on the NFM distance metric will achieve the least error probability as com-pared to those based on any other points on the feature lines. Furthermore, a theoretical investigation is provided that under certain assumption the NFL is approximately equivalent to the NFM when the dimension of the feature space is high. The experimental evaluations on both simulated and real-life benchmark data concur with all the theoretical investigations, as well as indicate that the NFM is ef-fective for the classification of the data with a Gaussian distribution or with a distribution that can be reasonably approximated by a Gaussian.
