Optimal Geometric Model Matching under Full 3D Perspective

doi:10.1006/cviu.1995.1028

Computer Vision and Image Understanding

Volume 61, Issue 3, May 1995, Pages 351-364

https://doi.org/10.1006/cviu.1995.1028 Get rights and content

Abstract

Model-based object recognition systems have rarely dealt directly with 3D perspective while matching models to images. The algorithms presented here use 3D pose recovery during matching to explicitly and quantitatively account for changes in model appearance associated with 3D perspective. These algorithms use random-start local search to find, with high probability, the globally optimal correspondence between model and image features in spaces containing over 2¹⁰⁰ possible matches. Three specific algorithms are compared on robot landmark recognition problems. A full-perspective algorithm uses the 3D pose algorithm in all stages of search while two hybrid algorithms use a computationally less demanding weak-perspective procedure to rank alternative matches and updates 3D pose only when moving to a new match. These hybrids successfully solve problems involving perspective, and in less time than required by the full-perspective algorithm.

References (0)

Cited by (28)

An application independent review of multimodal 3D registration methods
2020, Computers and Graphics (Pergamon)
Citation Excerpt :
One more limitation of these methods is their heavy computation cost. An early solution to the 2D/3D registration problem is proposed by Beveridge [163], where a random-start local search procedure is used to arrive at a local optimum. The method uses a hybrid pose estimation algorithm with both full-perspective and weak-perspective camera models.
Registration is a ubiquitous operation in Visual Computing, with applications in 3D object retrieval among others. Registration is the process of overlaying two or more datasets taken from different viewpoints, at different times or by different sensors into a common reference frame. Multimodal registration is a special case where the data to be matched do not belong to the same modality and is challenging due to the diverse nature of the modalities involved which makes the creation of a distance function harder. Due to the large number of possible modality combinations and application fields, a considerable number of multimodal registration techniques have been proposed in diverse fields, including medicine and archaeology. This survey aims to unify 3D multimodal registration techniques (i.e. where at least one of the modalities is in 3D) across application domains, with the hope of providing an application-independent view and the potential for cross-fertilization. The problem of 3D multimodal registration is explicitly defined and the various methods are systematically categorized and described in terms of a number of important properties. Methods with publicly available source code have been compared on common datasets. A discussion on trends, observations and challenges for further research concludes the review.
A family of globally optimal branch-and-bound algorithms for 2D–3D correspondence-free registration
2019, Pattern Recognition
Citation Excerpt :
The assumption however is unreasonable in many scenarios, where an algorithm that is robust to high outlier rates is required. An early solution to 2D–3D registration from correspondence-free lines is proposed by Beveridge and Riseman [28] who use a local search procedure to iteratively arrive at local optima. They investigate how easy the problem is; evaluating expected run-time as a function of the number of lines and amount of clutter.
We present a family of methods for 2D–3D registration spanning both deterministic and non-deterministic branch-and-bound approaches. Critically, the methods exhibit invariance to the underlying scene primitives, enabling e.g. points and lines to be treated on an equivalent basis, potentially enabling a broader range of problems to be tackled while maximising available scene information, all scene primitives being simultaneously considered. Being a branch-and-bound based approach, the method furthermore enjoys intrinsic guarantees of global optimality; while branch-and-bound approaches have been employed in a number of computer vision contexts, the proposed method represents the first time that this strategy has been applied to the 2D–3D correspondence-free registration problem from points and lines. Within the proposed procedure, deterministic and probabilistic procedures serve to speed up the nested branch-and-bound search while maintaining optimality. Experimental evaluation with synthetic and real data indicates that the proposed approach significantly increases both accuracy and robustness compared to the state of the art.
Matching horizon features using a Messy Genetic Algorithm
2000, Computer Methods in Applied Mechanics and Engineering
Citation Excerpt :
We have previously demonstrated the feasibility of matching horizon lines extracted from digital imagery to horizons extracted from rendered terrain maps using a Random Starts Local Search Algorithm [3]. We have also reported on local search algorithms for matching both 2D [1,5,20] and 3D [4] line segment models to imagery. The restricted viewing geometry associated with the orientation correction problem removes the need to treat this as a 3D matching problem.
Three line segment matching techniques are compared on horizon line matching problems. The application is to reduce camera angle uncertainty for outdoor robots. Each of the three algorithms seeks an optimal many-to-many match between a subset of 2D line segments extracted from a rendered terrain map and segments extracted from an image. Two algorithms are non-deterministic, with one performing Random Starts Local Search and the other using a Messy Genetic Algorithm. The third algorithm searches a list of ranked key feature matches using a limited deterministic local search. The Random Starts Local Search Algorithm is the slowest of the three algorithms and fails to find the best match on 9 out of 54 problems. In contrast, the Messy Genetic Algorithm correctly solves all 54 test problems. The Key Feature Algorithm is the fastest, but it fails to find the optimal match in three cases. The results suggest the Messy Genetic Algorithm is superior to the other two algorithms on problems of this type. The test imagery was collected on one of the Unmanned Ground Vehicle Program's vehicles operating at the Lockheed Martin Demo C test site. Using the optimal matches it is shown that uncertainty in camera pointing angle is reduced from several degrees to less than a degree.
Robust hypothesis verification: Application to model-based object recognition
1999, Pattern Recognition
The use of hypothesis verification is recurrent in the model based recognition literature. Small sets of features forming salient groups are paired with model features. Pose can be hypothesised from this small set of correspondences. Verification of the pose consists in measuring how much model features transformed by the computed pose coincide with image features. When data involved in the initial pairing are noisy the pose is inaccurate and verification is a difficult problem. In this paper we propose to use a robust hypothesis verification algorithm to perform object recognition. We explain how to integrate it in two different recognition schemes (2D and 3D recognition). After describing these applications we present numerous experimental results proving the robustness and the efficiency of these algorithms.
Solution of the Simultaneous Pose and Correspondence Problem Using Gaussian Error Model
1999, Computer Vision and Image Understanding
The use of hypothesis verification is recurrent in the model-based recognition literature. Verification consists in measuring how many model features transformed by a pose coincide with some image features. When data involved in the computation of the pose are noisy, the pose is inaccurate and difficult to verify, especially when the objects are partially occluded. To address this problem, the noise in image features is modeled by a Gaussian distribution. A probabilistic framework allows the evaluation of the probability of a matching, knowing that the pose belongs to a rectangular volume of the pose space. It involves quadratic programming, if the transformation is affine. This matching probability is used in an algorithm computing the best pose. It consists in a recursive multiresolution exploration of the pose space, discarding outliers in the match data while the search is progressing. Numerous experimental results are described. They consist of 2D and 3D recognition experiments using the proposed algorithm.
The Ascender System: Automated Site Modeling from Multiple Aerial Images
1998, Computer Vision and Image Understanding
The Ascender system acquires, extends, and refines 3D geometric site models from calibrated aerial imagery. To acquire a new site model, an automated building detector is run on one image to hypothesize potential building rooftops. Supporting evidence is located in other images via epipolar line segment matching in constrained search regions. The precise 3D shape and location of each building is then determined by multiimage triangulation under geometric constraints of 3D orthogonality, parallelness, colinearity, and coplanarity of lines and surfaces. Projective mapping of image intensity information onto these polyhedral building models results in a realistic site model that can be rendered using virtual “fly-through” graphics. As new images of the site become available, model extension and refinement procedures are performed to add previously unseen buildings and to improve the geometric accuracy of the existing 3D building models. In this way, the system gradually accumulates evidence over time to make the site model more complete and more accurate. An extensive performance evaluation of component algorithms and the full system has been carried out. Two-dimensional building detection accuracy, as well as accuracy of the three-dimensional building reconstruction, are presented for a representative data set.

View all citing articles on Scopus

View full text

Regular ArticleOptimal Geometric Model Matching under Full 3D Perspective

Abstract

Regular Article
Optimal Geometric Model Matching under Full 3D Perspective