Incremental point-based integration of registered multiple range images

Zhou, Hong; Liu, Y.

doi:10.1109/iecon.2005.1568950

Cited by 10 publications

(54 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Closest points from two registered views are obtained using dsearchn function of MATLAB. The points which have found correspondents are stored in S old-overlapping and S new-overlapping where as the points without correspondents is stored in S old-non-overlapping and S new-non-overlapping respectively [22]. Overlapping point sets are merged by evaluating average of each correspondent pairs of coordinates in S old-overlapping and S new-overlapping .…”

Section: Integrationmentioning

confidence: 99%

Feature Detection of an Object by Image Fusion

Pati¹,

Dutta²,

Barua³

2010

IJCA

View full text Add to dashboard Cite

In this paper, we propose a novel method for feature detection of an object by fusion of range and intensity images. For this purpose, we have developed a data acquisition system with a laser source and camera interfaced with Silicon Graphics machine. 3-D mesh representation of the surface of the object is obtained from the acquired range images. Extraction of structural features from the range images has been performed by two methods i.e. coordinate thresholding technique and Laplacian of Gaussian (LoG) edge detector. Extraction of structural features from the intensity image of the object has been performed by the Hough transform technique and Canny edge detector. An approach using shape signatures has been proposed to detect corner points in the edge maps obtained using LoG detector as well as Canny detector. The extracted 3-D edge maps as well as the detected corner points have been mapped to 2-D plane. The methodology for manual fusion of edge maps with the help of affine transformation and the concept of automatic fusion of edge maps by affine transformation followed by iterative closest point (ICP) algorithm have been introduced in this work. The automated technique for fusion overcomes the drawbacks associated with manual fusion. The fusion algorithm provides a composite image with more accurate and reliable information about the important features of the object.

show abstract

Section: Integrationmentioning

confidence: 99%

Feature Detection of an Object by Image Fusion

Pati¹,

Dutta²,

Barua³

2010

IJCA

View full text Add to dashboard Cite

show abstract

“…• 0D point-based integration method [23,24] directly detects and processes corresponding points in the registered overlapping range images. While the corresponding points in the new images are removed and recovered if necessary in Ref.…”

Section: Previous Workmentioning

confidence: 99%

“…[23], they are fused in Ref. [24], considering the confidence of their measurement. This kind of method is computationally efficient, since all processing is based only on points whose number is significantly smaller than that of the triangles used as primitives in the 2D mesh-based integration method [18][19][20].…”

Section: Previous Workmentioning

confidence: 99%

“…The 2nd, 6th, 10th, 14th, 18th, 22nd rows: the mesh-based method [19]. The 3rd, 7th, 11th, 15th, 19th, 23rd rows: the point-based method [24]. The 4th, 8th, 12th, 16th, 20th, 24th rows: the new clustering integration method.…”

Section: Previous Workmentioning

confidence: 99%

“…After having pre-processed the input data [24], a pair of point sets and their normal vectors is obtained. Then one point set and its normal vectors have to be transformed into the coordinate system of the other point set before integration.…”

Section: Clustering Integration Processmentioning

confidence: 99%

See 2 more Smart Citations