The problem of cloud tracking within a sequence of geo-stationary satellite images has direct relevance to the analysis of cloud life cycles and to the detection of cloud motion vectors (CMVs) The proposed approach first identifies a homogeneous consistent cloud mass for tracking and then establishes motion correspondence within an image sequence In contrast to the crosscorrelation based approach as adopted in automatic CMV detection analysis, a scale space classifier is designed to detect cloud mass in the source image taken at time t and the destination image at time t+/spl delta/t Boundaries of the extracted cloud segments are matched by computing a correspondence between high curvature points This shape based method is capable of tracking in the cases of rotation, scaling, and shearing, while the correlation technique is limited to translational motion The final tracking results provide motion magnitude and direction for each contour point, allowing reliable estimation of meteorological events and wind velocities aloft With comparable computational expense, the scale space classification technique exceeds the performance of the traditional correlation-based approach in terms of reduced localization error and false matches

/pdf/cloud-tracking-by-scale-space-classification-1t01krqyv8.pdf

Cloud tracking by scale space classification

Motion analysis is an important component of computer vision, dealing with the measurement of motion information from a time varying image sequence. The analysis of dynamic image sequences finds a wide range of applications, including robot navigation, target tracking, traffic monitoring, motion of biological cells, cloud and weather systems tracking, image coding, etc. The paper reviews the various techniques used in motion analysis from image sequences. The approaches are classified according to the methodology in two main groups: methods based on the optical flow constraint equation, and matching. The advantages and disadvantages of both methodologies are also discussed.

Review of motion analysis techniques

Attempts to concisely describe dynamic phenomena recorded by image sequences tend to model the depicted scene as a configuration of objects which exhibit well defined (sequences of) state transitions. Current research in image sequence analysis is concerned with the design of computer-internal representations for objects and associated dynamic phenomena.

Overview on Image Sequence Analysis

This paper describes a model-based approach to perform tracking of extratropical atmospheric disturbances from a sequence of satellite cloud-cover images. More precisely, it deals with the estimation of motion of these spiral-shaped cloud systems (both translational and rotational motion), and the measurement of the evolution of their shape. Tracking is achieved by recording from one image to the next the changes of the model parameter values. A maximum likelihood criterion is used in the process of fitting model to sensed data. The defined model takes into account geometric and intensity aspects. Such an approach readily yields global information on the disturbance cloud system of interest. As a requirement in such an application is robustness to noise, to this end two versions of the modeling have been considered.

Modeling of Atmospheric Disturbances in Meteorological Pictures

Des chroniques d'images sont utilisees pour la mise en evidence de phenomenes dynamiques. Trois applications ont ete abordees dans un cadre operationnel et ont permis de definir une methodologie generale dont les etapes sont : 1) une mise en coincidence geometrique des images, 2) un traitement des images pour l'obtention de luminances normalisees, 3) un filtrage specifique prenant en compte le caractere spatial ou temporel du processus etudie. La comparaison et le traitement automatique d'images representant une meme scene necessitent d'une part la mise en coincidence geometrique des images. Pour cela, differentes methodes de correlation entre images sont etablies puis comparees. D'autre part, les luminances doivent etre normalisees afin de corriger leurs variations au cours du temps. La normalisation est effectuee soit par un ajustement du codage sur la dynamique des valeurs, soit par une modelisation de l'evolution temporelle de la luminance en chaque point. La premiere etude concerne la detection et le suivi de phenomenes meteorologiques sur des images de la Terre, fournies par des satellites meteorologiques, et affichees sur un ecran couleur. On a developpe une methode d'animation de ces images en vue d'une aide a la prevision meteorologique. Si l'objet de la premiere etude residait dans le suivi des nuages, le but de la seconde est la construction d'une image sans nuage grâce a un filtrage adaptatif et iteratif, afin d'obtenir une image du sol vu par ciel clair. La troisieme etude concerne la detection precoce de fumees d'incendie de foret. Quelques experiences ont permis de definir des caracteristiques spatiales et temporelles des fumees qui ont fourni les criteres de detection.

https://pastel.archives-ouvertes.fr/pastel-00956228/document

Définition d'une méthodologie de détection de phénomènes dynamiques dans une chronique d'images et son application à quelques études

In this paper, we propose an extraction method of some features, which describe dynamic properties of atmospheric disturbances. We identify a disturbance in a satellite image by means of a spirals model, knowing that the luminance model is constant levels plus noise. Thus, we define a class-conditional log-likelihood functional, what leads to a least-squares procedure. We compute the estimation of the shape change and motion parameters with a gradient algorithm. Such a processing could be an aid in interactive treatments and in the cloud archives development.

P. Bouthemy

Papers

Atmospheric Disturbances Tracking in Satellite Images