Name
Abstract | ||
|---|---|---|
Abstract Aframe,decimation ,scheme ,is proposed ,that makes ,automatic ,extraction of Structure and ,Motion ,(SaM) from ,handheld ,sequences ,more ,practical. Decimation of the number,of frames used for the actual SaM calculations keeps the size of the problem manageable, regardless of the input frame rate. The proposed,preprocessor is based ,upon global motion estimation between ,frames and a sharpness measure. With these tools, shot boundary detection is first performed,followed by the removal of redundant frames. The frame decimation makes it feasible to feed the system with a high frame rate, which in turn avoids loss of connectivity ,due to matching ,difficulties. A high ,input frame rate also enables robust automatic detection of shot ,boundaries. The development ,of the preprocessor was prompted by experience with a number of test sequences, acquired directly from a handheld ,camera. The preprocessor was tested on this material together with a SaM algorithm. The scheme,is conceptually simple and still has clear benefits. 1,Introduction Recently, the Structure and Motion (SaM) branch of computer vision has matured enough,to shift some of the interest to building reliable and practical algorithms and systems. The context considered here is the task of recovering ,camera positions and structure seen in a large number ,of views ,of a ,video sequence. Special interest is devoted,to a ,system ,that processes video directly from ,an initially uncalibrated camera, to produce a three-dimensional graphical model completely automatically. Great advances have been made ,towards this goal and a number ,of algorithms ,have been developed [4,8,10,11,15,18,24,26]. However, several additional pieces are necessary for an algorithm ,to become ,a full working ,system and these issues have been relatively neglected in the literature. One such piece, which is proposed here, is a preprocessing mechanism,able to produce,a sparse but sufficient set of views suitable for SaM. This mechanism,has several benefits. The most important benefit is that the relatively expensive SaM processing can be performed on a smaller number,of views. Another benefit is that video sequences with different amounts ,of motion ,per frame |
Year | DOI | Venue |
|---|---|---|
2000 | 10.1007/3-540-45296-6_2 | 3D Structure from Multiple Images of Large-Scale Environments |
Keywords | Field | DocType |
frame decimation,actual sam calculation,proposed preprocessor,sam algorithm,automatic extraction,redundant frame,frame decimation scheme,input frame rate,high input frame rate,high frame rate,computer vision,graphical model | Computer vision,Decimation,Computer science,Mobile device,Boundary detection,Residual frame,Preprocessor,Frame rate,Artificial intelligence,Motion estimation | Conference |
ISBN | Citations | PageRank |
3-540-41845-8 | 24 | 1.56 |
References | Authors | |
29 | 1 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| David Nistér | 1 | 2265 | 118.02 |