The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
This paper presents a technique for visual recognition in which physical objects are represented by families of surfaces in a local appearance space. An orthogonal family of local appearance descriptors is obtained by applying principal components analysis to image neighborhoods. The principal components with the largest variance are used to define a space for describing local appearance. The projection...
The foreground group in a scene may be ‘discovered’ and computed as a factorized approximation to the pairwise affinity of the elements in the scene. A pointwise approximation of the pairwise affinity information may in fact be interpreted as a ‘saliency’ index, and the foreground of the scene may be obtained by thresholding it. An algorithm called ‘affinity factorization’ is thus obtained which may...
This paper addresses a problem arising in the reverse engineering of solid models from depth-maps. We wish to identify and fit surfaces of known type wherever these are a good fit. This paper presents a set of methods for the least-squares fitting of spheres, cylinders, cones and tori to three-dimensional point data. Least-squares fitting of surfaces other planes, even of simple geometric type, has...
In this contribution, we are concerned with the detection and refined localization of 3D point landmarks. We propose multi-step differential procedures for subvoxel localization of 3D point landmarks. Moreover, we address the problem of choosing an optimal size for a region-of-interest (ROI) around point landmarks. That is, to reliably localize the landmark position, on the one hand, as much as possible...
This paper describes a color region-based approach to motion estimation in color image sequences. The system is intended for robotic and vehicle guidance applications where the task is to detect and track moving objects in the scene. It belongs to the class of feature-based matching techniques and uses color regions, resulting from a prior color segmentation, as the matching primitives. In contrast...
This paper proposes an approach for automatic road extraction in aerial imagery which exploits the scale-space behavior of roads in combination with geometric constrained snake-based edge extraction. The approach not only has few parameters to be adjusted, but for the first time allows for a bridging of shadows and partially occluded areas using the heavily disturbed evidence in the image. The road...
Computer vision systems, such as “seeing” robots, aimed at functioning robustly in a natural environment rich on information benefit from relying on multiple cues. Then the problem of integrating these become central. Existing approaches to cue integration have typically been based on physical and mathematical models for each cue and used estimation and optimization methods to fuse the parameterizations...
We develop an approach to image segmentation for natural scenes containing image texture. One general methodology which shows promise for solving this problem is to characterize textured regions via their responses to a set of filters. However, this approach brings with it many open questions, including how to combine texture and intensity information into a common descriptor and how to deal with...
Condensation, recently introduced in the computer vision literature, is a particle filtering algorithm which represents a tracked object's state using an entire probability distribution. Clutter can cause the distribution to split temporarily into multiple peaks, each representing a different hypothesis about the object configuration. When measurements become unambiguous again, all but one peak, corresponding...
A number of vision-based biometric techniques have been proposed in the past for personal identification. We present a novel one based on visual capturing of signatures. This paper describes a system based on correlation and recursive prediction methods that can track the tip of the pen in real time, with sufficient spatio-temporal resolution and accuracy to enable signature verification. Several...
An autonomous vehicle has been developed for precision application of treatment on outdoor crops. This document details a new vision algorithm to aid navigation and crop/weed discrimination being developed for this machine. The algorithm tracks a model of the crop planting pattern through an image sequence using an extended Kalman filter. A parallel update scheme is used to provide not only navigation...
This contribution attempts to move beyond the status where single moving objects in video image sequences are tracked separately in the scene domain, based on individually adapted approaches and parameters. Instead, we investigate which performance can be achieved by a combination of approaches based on edge element orientation and on optical flow, applied to a variety of image sequences and vehicles...
We propose a model for view-based adaptive affine tracking of moving objects. We avoid the need for feature-based matching in establishing correspondences through learning landmarks. We use an effective bootstrapping process based on colour segmentation and selective attention. We recover affine parameters with dynamic updates to the eigenspace using most recent history and perform predictions in...
Tracking the 2D contour of a moving object has widely been used in the past years. So called active contour models have been proven to be a promising approach to real-time tracking of deformable objects. Also tracking 2D contours, which are projections of rigid 3D objects, is reduced to tracking deformable 2D contours. There, the deformations of the contour are caused by the movement in 3D and the...
Image-sequence analysis for real-time applications requires high quality and highly efficient algorithms for tracking as there is no time to do the costly object recognition each time a new image is captured. Tracking with projection histograms revealed amazing results compared with standard correlation methods. Trackers based on projection histograms performed 31% up to 211% better than the reference...
W4S is a real time visual surveillance system for detecting and tracking people and monitoring their activities in an outdoor environment by integrating realtime stereo computation into an intensitybased detection and tracking system. Unlike many systems for tracking people, W4S makes no use of color cues. Instead, W4S employs a...
Tracking research has diverged into two camps; low-level approaches which are typically fast and robust but provide little fine-scale information, and high-level approaches which track complex deformations in high-dimensional spaces but must trade off speed against robustness. Real-time high-level systems perform poorly in clutter and initialisation for most high-level systems is either performed...
The recognition of human gestures and facial expressions in image sequences is an important and challenging problem that enables a host of human-computer interaction applications. This paper describes a framework for incremental recognition of human motion that extends the “Condensation” algorithm proposed by Isard and Blake (ECCV'96). Human motions are modeled as temporal trajectories of some estimated...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.