• 検索結果がありません。

要 旨 三次元変形を受けた画像に対するマッチング手法の比較

N/A
N/A
Protected

Academic year: 2021

シェア "要 旨 三次元変形を受けた画像に対するマッチング手法の比較"

Copied!
2
0
0

読み込み中.... (全文を見る)

全文

(1)

要 旨

三次元変形を受けた画像に対するマッチング手法の比較

奥山 晃平

本研究では,空間的回転の影響を受けた画像からの物体の認識を行うための基礎的実 験を行う.例えば,Google Street Viewやライブカメラなどに代表されるような空間的 な変形を受けた画像によるサービスにおいて,その膨大なデーターベースの中から効 率的に有用な情報を探すことは重要である.本研究では,三次元変形を受けた画像に対 してSIFT(Scale-Invariant Feature Transform)SURF(Speeded Up Robust Features) KAZEASIFT(Affine-SIFT)を用い,正答率の比較を行う.実験では,奥行き方向に3D 回転した画像を用いる実験とGoogle Street Viewより取得した画像を用いる実験の二種類 の実験を行う.奥行き方向に3D回転した画像を用いる実験では,画像を30度,60度,70 度,80度に回転したものと元画像のマッチングを40枚の画像に対して行う.Google Street Viewの画像の実験では,画像の取得にGoogle Street View Image API を用いる.取得す る画像は,ある地点の画像と,同じ地点から座標,カメラの方位,画像の水平視野を変更し た画像とし,100枚の画像に対してマッチングを行う.実験の結果から,奥行き方向に3D 回転した画像においてはASIFTが,Google Street Viewの画像に対してはSIFTKAZE が有効であることを確認している.

キーワード 三次元変形,局所特徴,SIFTSURFKAZEASIFT

i

(2)

Abstract

Comparison of Local Features applying 3-D transformed Image

Okuyama Kouhei

The aim of this research is to give a quantitative comparison of local features in or- der to make a matching between 3-D transformed imagesRecently three are many huge image databases and some of them are spatial image data such as Google Street View When we extract useful information from those imagewe often use image matching Local features are widely used for image matchingLocal features are designed for 2-D transformed images and their robustness for 3-D transformed images are not clearly un- derstoodIn this researchSIFT(Scale-Invariant Feature Transform)SURF(Speeded Up Robust Features)KAZEASIFT(Affine-SIFT) are applied to 3-D transformed im- ages to match the imagesThree dimensionalAffine transformed data(the angles are 30607080) and the images retrieved from Google Street View with different an- gleposition parameters are usedThe result shows that ASIFT is the best for Affine transformed images and SIFTKAZE are suitable to Google Street View images

key words 3-D transfomationComputer VisionScale-Invariant Feature Trans- formSpeeded Up Robust FeaturesKAZEAffine-SIFT

ii

参照

関連したドキュメント

In this case, the extension from a local solution u to a solution in an arbitrary interval [0, T ] is carried out by keeping control of the norm ku(T )k sN with the use of

Keywords: continuous time random walk, Brownian motion, collision time, skew Young tableaux, tandem queue.. AMS 2000 Subject Classification: Primary:

This article concerns the behaviour of solutions to a coupled sys- tem of Schr¨ odinger equations that has applications in many physical problems, especially in nonlinear optics..

Thus, we use the results both to prove existence and uniqueness of exponentially asymptotically stable periodic orbits and to determine a part of their basin of attraction.. Let

Section 4 will be devoted to approximation results which allow us to overcome the difficulties which arise on time derivatives while in Section 5, we look at, as an application of

In this paper, this problem will be solved for the case N = 2, for tested convex sets of class C 4 and testing convex sets of class C 2 , as stated in Theorem 2.2 below. From now on,

Section 3 is first devoted to the study of a-priori bounds for positive solutions to problem (D) and then to prove our main theorem by using Leray Schauder degree arguments.. To show

This paper presents an investigation into the mechanics of this specific problem and develops an analytical approach that accounts for the effects of geometrical and material data on