Sylvia Gil

Attentive mechanisms for dynamic and static scene analysis

Attention mechanisms extract regions of interest from image data to reduce the amount of information to be analyzed by time-consuming processes such as image transmission, robot navigation, and object recognition. Two such mechanisms are described. The first one is an alerting system that extracts moving objects in a sequence through the use of multiresolution representations. The second one detects regions in still images that are likely to contain objects of interest. Two types of cues are used and integrated to compute the measure of interest. First, bottom-up cues result from the decomposition of the input image into a number of feature and conspicuity maps. The second type of cues is top-down, and is obtained from a priori knowledge about target objects, represented through invariant models. Results are reported for both the alerting and the attention mechanisms using cluttered and noisy scenes.

COMPARING FEATURES FOR TARGET TRACKING IN TRAFFIC SCENES

This paper describes a motion-analysis system, applied to the problem of vehicle tracking in real-world highway scenes. In a first stage a motion-detection algorithm performs a figure/ground segmentation, providing binary masks of the moving objects. In the second stage, vehicles are tracked by using Kalman filters for two state vectors, which represent each targets position and velocity. Three types of features have been used: (i) the bounding rectangle, (ii) the centroid of the convex polygon approximating the vehicles contour and (iii) the 2-D pattern of the vehicle. For each feature, the performance of the tracking algorithm has been tested in terms of robustness and computing time.

Combining Multiple Motion Estimates for Vehicle Tracking

In this paper, the problem of combining estimates provided by multiple models is considered, with application to vehicle tracking. Two tracking systems, based on the bounding-box and on the 2-D pattern of the targets, provide individual motion parameters estimates to the combining method, which in turn produces a global estimate. Two methods are proposed to combine the estimates of these tracking systems: one is based on their covariance matrix, while the other one employs a Kalman filter model. Results are provided on three image sequences taken under different viewpoints, weather conditions and varying vehicle/road contrasts. Two evaluations are made. First, the performances of individual and global estimates are compared. Second, the two global estimates are compared and the superiority of the second method is assessed over the first one.

Computer vision and agricultural robotics for disease control: the Potato Operation

The Potato Operation is a project belonging to the domain of agricultural robotics. It aims to automate pulp sampling of potatoes in order to detect viral diseases. The results of the analysis of these samples are used for the certification of potatoes marked out as seed. The difficulty of this problem lies in the high variability of natural objects, such as their shape, colour or texture. This paper presents two computer vision approaches that have been implemented and tested, as well as the robotic apparatus required for the complete installation. The first computer vision approach, the so-called 3-D image analysis, is based on a combination of classical image analysis methods with active vision. The second approach, the so-called thermography, combines analysis of thermal images of potatoes with active vision. Other methods are also briefly described, which could lead to possible improvements of the results; they rely on alternative lighting schemes, such as fluorometry or the use of multiple sources. Finally, results of the different methods are presented and the long-term goals of the project are discussed.

FEATURE SELECTION FOR OBJECT TRACKING IN TRAFFIC SCENES

This paper describes a motion-analysis system, applied to the problem of vehicle tracking in real-world highway scenes. The system is structured in two stages. In the first one, a motion- detection algorithm performs a figure/ground segmentation, providing binary masks of the moving objects. In the second stage, vehicles are tracked for the rest of the sequence, by using Kalman filters on two state vectors, which represent each targets position and velocity. A vehicles motion is represented by an affine model, taking into account translations and scale changes. Three types of features have been used for the vehicles description state vectors. Two of them are contour-based: the bounding box and the centroid of the convex polygon approximating the vehicles contour. The third one is region-based and consists of the 2-D pattern of the vehicle in the image. For each of these features, the performance of the tracking algorithm has been tested, in terms of the position error, stability of the estimated motion parameters, trace of the motion models covariance matrix, as well as computing time. A comparison of these results appears in favor of the use of the bounding box features.

3D Image analysis for agricultural robotics: the Potato Operation

Abstract: The Potato Operation aims to automate pulp sampling in potatoes in order to detect viral activity. This article presents the vision aspect of the project. The goal is to compute the 3D coordinates of the biggest germ. This is accomplished by means of active vision and image analysis methods. A test system has been installed and experiments have been performed on a large sample, yielding a low error rate.

Potato operation: computer vision for agricultural robotics

Each year at harvest time millions of seed potatoes are checked for the presence of viruses by means of an Elisa test. The Potato Operation aims at automatizing the potato manipulation and pulp sampling procedure, starting from bunches of harvested potatoes and ending with the deposit of potato pulp into Elisa containers. Automatizing these manipulations addresses several issues, linking robotic and computer vision. The paper reports on the current status of this project. It first summarizes the robotic aspects, which consist of locating a potato in a bunch, grasping it, positioning it into the camera field of view, pumping the pulp sample and depositing it into a container. The computer vision aspects are then detailed. They concern locating particular potatoes in a bunch and finding the position of the best germ where the drill has to sample the pulp. The emphasis is put on the germ location problem. A general overview of the approach is given, which combines the processing of both frontal and silhouette views of the potato, together with movements of the robot arm (active vision). Frontal and silhouette analysis algorithms are then presented. Results are shown that confirm the feasibility of the approach.