A Region-Based Approach to Digital Image Registration with Subpixel Accuracy

Goshtasby, A. Ardeshir; Stockman, George C.; Page, Carl V.

doi:10.1109/tgrs.1986.289597

Cited by 188 publications

(79 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These range from relaxation-based methods [19,4], to cluster detection in transformation space (by computing point-to-point correspondences [20][21][22], to hierarchical decomposition of transformation space coupled with the application of a robust similarity measure [2,11,23,24]. Most of the techniques presented in these papers are computationally intensive (in a worst-case theoretical sense), or take long times to run in practice.…”

Section: Prior Workmentioning

confidence: 99%

“…Thus, by sampling a number of triples of points (or even pairs or singletons, depending on the number of degrees of freedom in the transformation space), and enumerating all possible matching triples in the other point set, one is likely to encounter a good match. This is the basis of many matching algorithms, including those of Goshtasby and others [20][21][22] and Goodrich et al [17]. and also of methods based on geometric hashing [25,26].…”

Section: Prior Workmentioning

confidence: 99%

See 1 more Smart Citation

Efficient algorithms for robust feature matching

Mount

Netanyahu

Moigne

1999

Pattern Recognition

135

103

View full text Add to dashboard Cite

One of the basic building blocks in any point-based registration scheme involves matching feature points that are extracted from a sensed image to their counterparts in a reference image. This leads to the fundamental problem of point matching: Given two sets of points, find the (affine) transformation that transforms one point set so that its distance from the other point set is minimized. Because of measurement errors and the presence of outlying data points, it is important that the distance measure between the two point sets be robust to these effects. We measure distances using the partial Hausdorff distance. Point matching can be a computationally intensive task, and a number of theoretical and applied approaches have been proposed for solving this problem. In this paper, we present two algorithmic approaches to the point matching problem, in an attempt to reduce its computational complexity, while still providing a guarantee of the quality of the final match. Our first method is an approximation algorithm, which is loosely based on a branch-andbound approach due to Huttenlocher and Rucklidge, (Technical Report 1321, Dept. of Computer Science, Cornell University, Ithaca, 1992; Proc. IEEE Conf. on Computer vision and Pattern Recognition, New York, 1993, pp. 705-706). We show that by varying the approximation error bounds, it is possible to achieve a tradeoff between the quality of the match and the running time of the algorithm. Our second method involves a Monte Carlo method for accelerating the search process used in the first algorithm. This algorithm operates within the framework of a branch-and-bound procedure, but employs point-to-point alignments to accelerate the search. We show that this combination retains many of the strengths of branch-and-bound search, but provides significantly faster search times by exploiting alignments. With high probability, this method succeeds in finding an approximately optimal match. We demonstrate the algorithms' performances on both synthetically generated data points and actual satellite images.

show abstract

Section: Prior Workmentioning

confidence: 99%

Section: Prior Workmentioning

confidence: 99%

Efficient algorithms for robust feature matching

Mount

Netanyahu

Moigne

1999

Pattern Recognition

135

103

View full text Add to dashboard Cite

show abstract

“…We initialize a and b at 1 and 0, respectively, then update them using the partial differential equation of E with respect to a and b as shown in Eqs. (19) and (20). a and b can be updated simultaneously with the linear and nonlinear transformation optimization mentioned above.…”

Section: Original Articlementioning

confidence: 99%

“…Image registration algorithms can be categorized into linear and nonlinear approaches based on the mapping functions between reference and target images. Linear image registration approaches (20)(21)(22)(23) assume that the objects in the images are rigid. Their mapping functions include similarity transformation, affine transformation and perspective projection.…”

mentioning

confidence: 99%

3D image stack reconstruction in live cell microscopy of Drosophila muscles and its validation

Wasser

2009

Cytometry Pt A

View full text Add to dashboard Cite

Rapid movements of live tissues during the acquisition of 3D image stacks can result in misalignments between successive image slices. The remodeling of the muscles in Drosophila metamorphosis is an example where sporadic motion during image acquisition impede image analysis and volume visualization. Most of the image stack registration algorithms applied in microscopy are aimed at the linear alignment of fixed histological sections. However, live muscles are nonrigid objects and their contractions and relaxations represent nonlinear transformations that cannot be properly rectified by applying purely linear registration methods. We developed a fully automated area-based nonrigid stack registration (NSR) method that minimizes the mean square error of intensities between successive image slices. The mapping function is formulated using the thin plate spline (TPS). A hierarchical linear to nonlinear, coarse to fine matching strategy is applied to ensure stability and fast convergence. Topological structure is preserved by constraining the step size of the nonlinear transformation. To assess the accuracy of 3D reconstruction, we propose a new benchmarking method that measures geometrical features of restored nuclei. We tested our algorithm on image stacks generated by laser scanning confocal microscopy that show live muscles during the prepupal stage of Drosophila metamorphosis. Our registration algorithm is able to restore image stacks that are distorted by periodic contraction of muscles. Quantitative assessment of registration performance agrees well with qualitative visual inspection. Our NSR method is able to restore image stacks for the purpose of visualization and quantitative analysis of Drosophila metamorphosis and, potentially, various other processes in developmental biology studied by 3D live cell microscopy. ' International Society for Advancement of CytometryKey terms live cell imaging; confocal microscopy; 3D reconstruction; nonrigid registration; thin plate spline; Drosophila; muscle development; metamorphosis IN the areas of developmental and cell biology, multidimensional imaging of living tissues has gained increasing popularity as a tool to study and visualize the diversity of cellular functions and behaviors. The progress in live cell imaging is driven by two principal technologies: the development of fluorescent proteins and improvement in instrumentation. Ever since the discovery of the green fluorescent protein (GFP) from the jellyfish Aequorea (1), various studies have demonstrated that GFP, its variants and many other fluorophores from other (mostly marine) animals can be functionally tagged to almost any other protein (2,3). Through genetic engineering it is possible to control multiple aspects of reporter gene expression, including cell type, sub-cellular localization, time and strength, and the color of the fluorophore.Confocal laser scanning microscopy (CLSM) is the most common imaging modality to visualize fluorescently labeled cells in animal models, such as zebrafish, Caen...

show abstract

“…These indices are consistent with the definitions in Eqs. (4) and (5). The transformation TR, k can be used to transform the region in Figure 1 to the corresponding region in the reference coordinate frame.…”

Section: Registration Algorithm For Known Transformationmentioning

confidence: 99%

Algorithm for image registration and clutter and jitter noise reduction

Brower¹

1997

View full text Add to dashboard Cite

A Region-Based Approach to Digital Image Registration with Subpixel Accuracy

Cited by 188 publications

References 11 publications

Efficient algorithms for robust feature matching

Efficient algorithms for robust feature matching

3D image stack reconstruction in live cell microscopy of Drosophila muscles and its validation

Algorithm for image registration and clutter and jitter noise reduction

Contact Info

Product

Resources

About