Planar homography: accuracy analysis and applications
Summary (2 min read)
Introduction
- The registration of various frames in a video mosaic has numerous applications, including generation of terrain mosaics from flyover image transects in underwater and airborne systems [3].
- Claimed to be based on the earlier results of Weng et al. [7], who reported a comprehensive error analysis of motion and structure parameters from image correspondences.
- It is typically the case that some robust estimation method, e.g. RANSAC or LMedS, can be used as a first step to identify the outliers, before the proposed linear solution is applied to the inliers (which satisfy the assumed noise model).
3. COVARIANCE OF PROJECTIVE HOMOGRAPHY
- Due to space limitation, the authors refer the reader to section 5 in [4], given for the estimation of the covariance of the homography parameters.
- The authors also derive Cho, the covariance derived here as the new estimate.
- From δm, the authors can determine the variation in the measurement matrix A2N×9, or Q9×9.
4. SIMULATIONS
- The authors use computer simulations to compare the closed-form expressions for estimating the variances of the projective homography parameters, given in [4] and derived here (denoted Chc and Cho, respectively).
- In these two their simulations, only a minimum 4 points near the image corners are used.
- Measurements noise from a normal distribution with standard deviation σm is then added to corresponding pairs to subsequently estimate the homography, and calculate its error.
- The process is repeated 1000 times with different noise samples to compute statistical measures.
- The final experiment deals with the generation of a 10-frame sequence, and the integration of frame-to-frame homographies and the corresponding variances.
4.1. Case 1
- Blue crosses depict the errors of the homography parameters for each of 1000 simulations, with red circles giving the (zero) mean error.
- In [4], the authors claim that their solution provides better estimates in two cases: 1) Relatively small measurement noise levels, or 2) when minimum N = 4 image correspondences are utilized in the estimation of the homography.
- In all cases, the new results consistently provide a more accurate estimate of the ±3σh error bounds.
- Authorized licensed use limited to: UNIVERSITAT DE GIRONA.
4.2. Case 2
- Fig. 2 shows the original and transformed images based on an assumed homography.
- In addition, selected corners have been mapped with 1000 homographies, estimated from noisy correspondences (σm = 1 is assumed).
- The green and red uncertainty ellipses of the transformed points have been determined from the homography variances σhc and σho, respectively.
- The results for 4 selected points, A − D, demonstrate once again that σho provides a more accurate and tighter uncertainty bound.
4.3. Case 3
- A 10-frame sequence has been constructed based on a prescribed homography.
- Fig. 3 shows the sequence with an inter-frame motion of about 13-14 pixels.
- The blue stars depict the true positions of 3 sample pixels in various frames, at the center of uncertainty ellipses (computed from the two different techniques) that bound noisy positions of these points based on homographies estimated from 1000 simulations with noisy correspondences.
- For each of these three points, the noisy positions and bounding ellipses in frames 1, 4, 7 and 10 have been given in subsequent plots.
- As in previous 2 cases, their new results provide a tighter fit of the projected point distributions.
5. SUMMARY
- Ability to not only estimate the transformation between frames but also to assess the confidence in these estimates is important in many applications involving motion estimation from video imagery.
- Based on earlier results of Weng et al. [7], the authors have provided new expressions to estimate these bounds more accurately.
- These results are particularly important when dealing with long image sequences where frame-to-frame estimates need to be integrated to establish the camera position, to build an image mosaic, or generally to register various frames of a video sequence.
- Under investigation is the direct use of these uncertainty bounds in the construction of photo-mosaics.
- This work has been funded in part by the Spanish Ministry of Education and Science under grant CTM2004-04205, and in part by the Generalitat de Catalunya under grant 2003PIVB00032.
6. REFERENCES
- Authorized licensed use limited to: UNIVERSITAT DE GIRONA.
- Z. Sun, V. Ramesh, and A.M. Tekalp, “Error characterization of the factorization method,” CVIU, vol.
Did you find this useful? Give us your feedback
Citations
2 citations
1 citations
1 citations
References
15,558 citations
495 citations
"Planar homography: accuracy analysi..." refers background in this paper
...For small variations –max{δqi} << 1, where qi denotes i-th element of q –it can be shown [6, 7] that up to first-order, the eigenvalues and eigenvectors ofQ vary according to δλi = υi ∆Q υi and δυi = VΨiV T Πiδq where Ψi = diag{(λi − λ1)−1, ....
[...]
...[7], who reported a comprehensive error analysis of motion and structure parameters from image correspondences....
[...]
280 citations
265 citations
"Planar homography: accuracy analysi..." refers background or methods in this paper
...In addition to many methods for estimating the homography parameters [5], analytical expressions to assess the accuracy of the transformation parameters have been proposed [4]....
[...]
...[4] give closed-form expressions to estimate the variance of the 8 independent parameters1....
[...]
...SIMULATIONS We use computer simulations to compare the closed-form expressions for estimating the variances of the projective homography parameters, given in [4] and derived here (denoted Chc and Cho, respectively)....
[...]
...COVARIANCE OF PROJECTIVE HOMOGRAPHY Due to space limitation, we refer the reader to section 5 in [4], given for the estimation of the covariance of the homography parameters....
[...]
41 citations
Related Papers (5)
Frequently Asked Questions (7)
Q2. In what case does the author claim that their solution provides better estimates in two cases?
In [4], the authors claim that their solution provides better estimates in two cases: 1) Relatively small measurement noise levels, or 2) when minimum N = 4 image correspondences are utilized in the estimation of the homography.
Q3. What are the limitations of the analysis?
Computation of projective homography from frame-to-frame correspondences has been extensively studied in recent years [5], and analytical uncertainty bounds of the homography parameters and reprojection errors have been proposed [4].
Q4. What is the error bound for the envelops?
The dashed blue envelop is the ±3σ error bound computed experimentally, and the other two envelops in green and red are derived from analytical bounds ±3σhc and ±3σho, respectively.
Q5. What is the covariance of the homography?
The authors construct matching pairs {p, p′} based on a pre-specified homographyH; the authors use the well-know interpretation H = R + tnT in terms of the motion {R, t} of a camera relative to a planar scene with surface normal n = [−P,−Q, 1]/Zo, where P and Q control the surface slant and tilt angles, and Zo its distance from the camera.
Q6. How do the authors determine the covariance of the homography parameters?
For small variations –max{δqi} << 1, where qi denotes i-th element of q –it can be shown [6, 7] that up to first-order, the eigenvalues and eigenvectors ofQ vary according to δλi =
Q7. What is the purpose of this paper?
Ability to not only estimate the transformation between frames but also to assess the confidence in these estimates is important in many applications involving motion estimation from video imagery.