In this paper we propose to simultaneously detect lane and pavement boundaries by fusing information from both optical and radar images. The boundaries are described with concentric circular models, whose parameters are compatible and will result in better conditioned estimation problems than previous parabolic models. The optical and radar imaging processes are represented with Gaussian and log-normal probability densities, with which we successfully avoid the ad hoc weighting scheme carried on the two likelihood functions. The multisensor fusion boundary detection problem is posed in a Bayesian framework and a joint maximum a posteriori (MAP) estimate is employed to locate the lane and pavement boundaries. Experimental results have shown that the fusion algorithm outperforms single sensor based boundary detection algorithms in a variety of road scenarios. And it also yields better boundary detection results than the fusion algorithm that took advantage of existing prior and likelihood formulations.

Treats a problem arising in the design of intelligent vehicles: automated detection of lane and pavement boundaries using forward-looking optical and radar imaging sensors mounted on an automobile. In previous work, lane and pavement boundaries have always been located separately. This separate detection strategy is problematic in situations when either the optical or the radar image is too noisy. We propose a Bayesian multisensor image fusion method to solve our boundary detection problem. This method makes use of a deformable template model to globally describe the boundaries of interest. The optical and radar imaging processes are described with random field likelihoods. The multisensor fusion boundary detection problem is reformulated as a joint MAP estimation problem. However, the joint MAP estimate is intractable, as it involves the computation of a notoriously difficult normalization constant, also known as the partition function. Therefore, we settle for the so-called empirical MAP estimate, as an approximation to the true MAP estimate. Several experimental results are provided to demonstrate the efficacy of the empirical MAP estimation method in simultaneously detecting lane and pavement boundaries. Fusion of multi-modal images is not only of interest to the intelligent vehicles community, but to others as well, such as biomedicine, remote sensing, target recognition. The method presented in the paper is also applicable to image fusion problems in these other areas.

This paper introduces a new algorithm for detecting lane markers in images acquired from a forward-looking vehicle-mounted camera. The method is based on a novel set of frequency domain features that capture relevant information concerning the strength and orientation of spatial edges. The frequency domain features are combined with a deformable template prior, in order to detect the lane markers of interest. Experimental results that illustrate the performance of this algorithm on images with varying lighting and environmental conditions, shadowing, lane occlusion(s), solid and dashed lines, etc. are presented. The algorithm seems to detect lane markers remarkably well under a very large and varied collection of roadway images. An experimental comparison is drawn between this frequency feature-based algorithm and the spatial feature-based LOIS lane detection algorithm.

This paper treats automated detection of road and lane boundaries by fusing information from forward-looking optical and active W-band radar imaging sensors mounted on a motor vehicle. A deformable template model is used to globally describe the boundary shapes. The optical and radar imaging processes are characterized with random field likelihoods. The multisensor fusion edge detection problem is posed in a Bayesian framework and a joint MAP estimate is employed to locate the road and lane boundaries. Three optimization approaches (multi-resolution pseudo-exhaustive search, Metropolis algorithm, and Metropolis algorithm with pre-tuned curvature) are proposed to implement the joint MAT estimate. Experimental results are shown to demonstrate that the joint MAP algorithm operates robustly and efficiently in a variety of road scenarios.

This paper introduces a new algorithm, called lane-finding in another domain (LANA), for detecting lane markers in images acquired from a forward-looking vehicle-mounted camera. The method is based on a novel set of frequency domain features that capture relevant information concerning the strength and orientation of spatial edges. The frequency domain features are combined with a deformable template prior, in order to detect the lane markers of interest. Experimental results that illustrate the performance of this algorithm on images with varying lighting and environmental conditions, shadowing, lane occlusion(s), solid and dashed lines, etc. are presented. LANA detects lane markers well under a very large and varied collection of roadway images. A comparison is drawn between this frequency feature-based LANA algorithm and the spatial feature-based LOIS lane detection algorithm. This comparison is made from experimental, computational and methodological standpoints.

Three methods of detecting road edges in millimeter-wave radar images are presented. All of them are based on deformable template priors and random field likelihoods. The first method is formulated in a Bayesian setting and employs an adaptive MAP estimate. The second method is a modification of the first, using a novel weighting scheme. The third method is based on a three-region indicator matrix which is used to impose the non-linear constraints implicit on road geometry via addition of a sum of quasi-quadratic matrix forms to the log-normal likelihood. Unlike the first two methods, that employ the Metropolis algorithm to find the optimal road edges, the third method uses a deterministic recursive scheme designed to find the optimal indicator matrix. Experimental results are presented to show the advantages of these methods.

This paper proposes a segmentation algorithm using deformable template models to segment a vehicle of interest both from the stationary complex background and other moving vehicles in an image sequence. We define a polygonal template to characterize a general model of a vehicle and derive a prior probability density function to constrain the template to be deformed within a set of allowed shapes. We propose a likelihood probability density function which combines motion information and edge directionality to ensure that the deformable template is contained within the moving areas in the image and its boundary coincides with strong edges with the same orientation in the image. The segmentation problem is reduced to a minimization problem and solved by the Metropolis algorithm. The system was successfully tested on 405 image sequences containing multiple moving vehicles on a highway.

We propose a general object localization and retrieval scheme based on object shape using deformable templates. Prior knowledge of an object shape is described by a prototype template which consists of the representative contour/edges, and a set of probabilistic deformation transformations on the template. A Bayesian scheme, which is based on this prior knowledge and the edge information in the input image, is employed to find a match between the deformed template and objects in the image. Computational efficiency is achieved via a coarse-to-fine implementation of the matching algorithm. Our method has been applied to retrieve objects with a variety of shapes from images with complex background. The proposed scheme is invariant to location, rotation, and moderate scale changes of the template.

This paper addresses the problem of locating two straight and parallel road edges in images that are acquired from a stationary millimeter-wave radar platform positioned near ground-level. A fast, robust, and completely data-driven Bayesian solution to this problem is developed, and it has applications in automotive vision enhancement. The method employed in this paper makes use of a deformable template model of the expected road edges, a two-parameter log-normal model of the ground-level millimeter-wave (GLEM) radar imaging process, a maximum a posteriori (MAP) formulation of the straight edge detection problem, and a Monte Carlo algorithm to maximize the posterior density. Experimental results are presented by applying the method on GLEM radar images of actual roads. The performance of the method is assessed against ground truth for a variety of road scenes.

The paper addresses the problem of detecting lane boundaries in color images of road scenes acquired from a car mounted visual sensor. It is shown that the lane boundaries in such images have to obey a set of global constraint equations. All images with such constrained lanes are modeled via deformable templates. The observed image is related to the underlying lane boundary features through a likelihood function which is based on the degree of match (in magnitude/direction) between the deformed template and the lane edges. The lane detection problem is formulated in a Bayesian setting, and it is posed as an equivalent problem of maximizing a posterior pdf which sits over a low-dimensional deformation space. This pdf is multi-modal hence a Metropolis algorithm is employed to obtain its maximum. Experimental results are shown to illustrate the performance of this algorithm.

This paper presents two new methods for detecting edges in millimeter-wave radar images. The first method is based on a deformable template model of edge shapes and a random field model of the millimeter-wave imaging process. The second method is similar to the first, except that the imaging model component is replaced by one that is based on the magnitude and direction of the image gradient. Experimental results are shown to illustrate the advantages of using these methods over traditional edge detectors.

This paper presents a new method for representing the spatial information present in digital grey-tone images. The method is based on using multi-resolution decompositions (MRDs) and Markov random fields (MRFs) concurrently. A given image is represented by a MRD of it, along with an optimally estimated set of Gaussian MRF (GMRF) parameters. Since the GMRF parameters are very small in number, this addition to the usual MRD results in only a small increase in the number of bits in the representation. It is shown, however, that such a minor addition helps when reconstructing the (given) original image from its MRD. Experimental results are presented to illustrate the usefulness of this new method.>

The valid parameter spaces of infinite- and finite-lattice (2-D noncausal) Gaussian Markov random fields (GMRFs) are investigated. For the infinite-lattice fields, the valid parameter space is shown to admit an explicit description; a procedure that yields the valid parameter space is presented. This procedure is applied to the second-order (neighborhood) 2-D GMRFs to obtain an explicit description of their valid parameter spaces. For the finite-lattice fields, it is shown that the valid parameter space does not admit a simple description; the conditions that ensure the positivity of the power spectrum are necessary, sufficient, and irreducible. The set of conditions for the infinite-lattice fields, however, serves as a good set of sufficient conditions for the finite-lattice case. The results readily extend to the class of d-D real and complex GMRFs.>

An adaptive segmentation algorithm is developed which simultaneously estimates the parameters of the underlying Gibbs random field (GRF)and segments the noisy image corrupted by additive independent Gaussian noise. The algorithm, which aims at obtaining the maximum a posteriori (MAP) segmentation is a simulated annealing algorithm that is interrupted at regular intervals for estimating the GRF parameters. Maximum-likelihood (ML) estimates of the parameters based on the current segmentation are used to obtain the next segmentation. It is proven that the parameter estimates and the segmentations converge in distribution to the ML estimate of the parameters and the MAP segmentation with those parameter estimates, respectively. Due to computational difficulties, however, only an approximate version of the algorithm is implemented. The approximate algorithm is applied on several two- and four-region images with different noise levels and with first-order and second-order neighborhoods.>