본 논문에서는 마커 없이 증강 현실을 구현하기 위한 물체 인식 기법을 제안한다. 먼저 SIFT(Scale Invariant Feature Transform)알고리즘을 사용하여 물체 영상으로부터 특징점을 찾는데, 이러한 특징점들은 비율, 회전 또는 이동시에도 그 특징이 변하지 않는 장점이 있다. 또한 조도의 변화에도 일부는 변화지 않는 특성을 갖는다. 추출된 특징점의 독립적인 특성을 이용해 화면내의 다른 이미지의 매칭 포인트를 찾을 수 있는데, 학습된 영상과 매칭이 이루어지면, 매칭된 점을 이용해 화면내의 물체를 찾는다. 본 논문에서는 장면의 첫 프레임에서 발생하는 템플릿 이미지와의 매칭을 통해 현재의 화면에서 물체를 인식하였다. 네 종류의 물체에 대해 인식 실험을 한 결과 제안한 방법이 우수한 성능을 갖는 것을 확인하였다.
본 논문에서는 마커 없이 증강 현실을 구현하기 위한 물체 인식 기법을 제안한다. 먼저 SIFT(Scale Invariant Feature Transform)알고리즘을 사용하여 물체 영상으로부터 특징점을 찾는데, 이러한 특징점들은 비율, 회전 또는 이동시에도 그 특징이 변하지 않는 장점이 있다. 또한 조도의 변화에도 일부는 변화지 않는 특성을 갖는다. 추출된 특징점의 독립적인 특성을 이용해 화면내의 다른 이미지의 매칭 포인트를 찾을 수 있는데, 학습된 영상과 매칭이 이루어지면, 매칭된 점을 이용해 화면내의 물체를 찾는다. 본 논문에서는 장면의 첫 프레임에서 발생하는 템플릿 이미지와의 매칭을 통해 현재의 화면에서 물체를 인식하였다. 네 종류의 물체에 대해 인식 실험을 한 결과 제안한 방법이 우수한 성능을 갖는 것을 확인하였다.
In this paper, we propose an object recognition technique for implementing marker less augmented reality. Scale Invariant Feature Transform (SIFT) is used for finding the local features from object images. These features are invariant to scale, rotation, translation, and partially invariant to illum...
In this paper, we propose an object recognition technique for implementing marker less augmented reality. Scale Invariant Feature Transform (SIFT) is used for finding the local features from object images. These features are invariant to scale, rotation, translation, and partially invariant to illumination changes. Extracted Features are distinct and have matched with different image features in the scene. If the trained image is properly matched, then it is expected to find object in scene. In this paper, an object is found from a scene by matching the template images that can be generated from the first frame of the scene. Experimental results of object recognition for 4 kinds of objects showed that the proposed technique has a good performance.
In this paper, we propose an object recognition technique for implementing marker less augmented reality. Scale Invariant Feature Transform (SIFT) is used for finding the local features from object images. These features are invariant to scale, rotation, translation, and partially invariant to illumination changes. Extracted Features are distinct and have matched with different image features in the scene. If the trained image is properly matched, then it is expected to find object in scene. In this paper, an object is found from a scene by matching the template images that can be generated from the first frame of the scene. Experimental results of object recognition for 4 kinds of objects showed that the proposed technique has a good performance.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
이론/모형
Many Researchers worked with natural object recognition without markers. The 2D feature extraction and matched with the priori known 3D models proposed by D.Beier [3] for markerless Augmented Reality Applications .Quan Wang et el.
Descriptors with lower dimension than this don't perform as well across the range of matching task. The computational cost remains low due to the approximate Best Bin First (BBF)method used for finding the nearest-neighbor. Best Bin First is an approximate algorithm which returns the nearest-neighbor for a large fraction of queries and a very close neighbor.
Many researchers worked with feature extraction. We used SIFT algorithm proposed by David Lowe [1]for extracting the features from images of objects. We first make some template images from different objects and find out the keypoints of those template images.
성능/효과
Best Bin First is an approximate algorithm which returns the nearest-neighbor for a large fraction of queries and a very close neighbor. To test the distinctiveness of the SIFT descriptors, matching accuracy is also measured against varying number of keypoints in the testing database, and it is shown that matching accuracy decreases only very slightly for very large database sizes.
참고문헌 (8)
David Ge, "Distinctive Image Features from Scale-Invariant Keypoints", Int. Journal of Computer Vision, Vol.60, No.2, pp.91-110, 2004.
David. G. Lowe, "Object recognition from local scale invariant features", Proc. of the Intnl. Canf. on Computer Vision, pp. 1150-1157, Corfu, Greece, 1999.
D. Beier, R. Billert, B. Bruderlin, D.Stichling, B. Kleinjohann, "Marker-less Vision Based Tracking for Mobile Augmented Reality" Proc. of IEEE Int. Conf.on Mixed and Augmented Realily, pp.258-259, Oct.2003.
Q. Wang and S. You ,"Real-Time image matching based on multiple view kernel projections" IEEE Conference on Computer Vision and Pattern Recognition, 2007.
Schiele, Bernt, and James L.Crowley," Object Recognition using multidimensional receptive field histograms", Fourth European Conference on Computer Vision,cambridge,UK(1996), pp.67-97.
Aibing Rao, and Rohini K. Srihari, "Spatial Color Histogramsfor Content-Based Image Retrieval", Proc. of IEEE Int. Conf. onTools with Artificial Intelligence, pp.183-186, Nov. 1999.
D. Marimon and T. Ebrahimi", Orientation histogram-based matching for region tracking" IEEE International workshop on Image Analysis and Multimedia interactive services, pp. 6-8 ,June 2007.
Schmid ,C. and R.Mohr,"Local grayvalue invariants for image retrieval", IEEE PAMI ,19,5,1997, pp530-534.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.