Image analysis – Applications – Target tracking or detecting
Reexamination Certificate
1999-06-07
2002-08-20
Johns, Andrew W. (Department: 2621)
Image analysis
Applications
Target tracking or detecting
Reexamination Certificate
active
06438253
ABSTRACT:
BACKGROUND AND SUMMARY OF THE INVENTION
This invention relates to a process for dynamic monitoring of changes to deformable media, and for predicting these changes.
In domains such as air traffic control, it is important not only to monitor trajectories followed by aircraft to guide them, but also to know the weather conditions within the radius of action of the instruments used for this control, and particularly for detecting and precisely localizing cloud formations that could be dangerous or simply unpleasant for persons in the aircraft. This is usually done using weather radars, for which the echo images are displayed on the control center screen after eliminating echoes due to fixed obstacles. Human interpretation of these types of image is often complicated and these images cannot be used to forecast changes to cloud formations. Furthermore, an air traffic controller needs to know the weather conditions and how they will change in order to optimize routing of aircraft that he is responsible for guiding and to understand any route changes decided upon by the captain to avoid cloud formations that he considers are dangerous. The article by F. BARBARESCO, S. BONNEY, J. LAMBERT and B. MONNIER published in IEEE International Conference on Image Processing—I.C.I.P. 96, in Lausanne (Switzerland) in September 1996 entitled “Motion Based Segmentation and Tracking of Dynamic Radar Clutter” describes a process for processing this type of image by determining active contours with a constraint model, but this process has two disadvantages; firstly it requires a prohibitive calculation time, and secondly it cannot manage complex deformations (it simply manages deformations that can be approximated by an affine deformation model).
The purpose of this invention is a process for dynamically monitoring the variation of deformable media that can be used for this monitoring and for creating relatively reliable forecasts of how the deformations will change, without requiring a prohibitive calculation time.
The process according to the invention for dynamic monitoring of deformable media consists of using at least two images taken at different times obtained by at least one sensor, establishing the skeleton of each distinct assembly in this medium for each image, and then making the image skeletons in question correspond to each other.
This correspondence is made by vectorizing the skeletons of the processed images, examining the vectors of these skeletons taken in pairs, and each time attempting to match two vectors with approximately the same geometric characteristics. Skeletons of assemblies, and if required the corresponding assemblies, can be rebuilt from these matched skeletons.
Advantageously, the image may be preprocessed before the skeletons are built up. This preprocessing may include a filter step (frequency and/or morphological filtering), and a thresholding step eliminating image components that are not relevant for the problem in question. The processing can advantageously be facilitated by simplifying the skeletons by vectorization (thinning and/or search for points at which curves representative of skeletons have a curvature greater than a determined value and then linearization) and/or elimination of insignificant artifacts or barbules.
According to one aspect of the process according to the invention, a map of displacement vectors is created for at least one area of interest in the processed images.
According to another aspect of the invention, a predicted image is built up starting from displacement vectors and/or the skeleton. Another possible way of rebuilding the above mentioned image is to use the skeleton displacement vectors to calculate the displacement of each pixel image and to rebuild the predicted image from the newly predicted pixel locations.
According to another aspect of the process according to the invention, attributes are applied to at least some of the pixels in several images taken at different times, the characteristics of the corresponding areas are determined, and information about the variation and/or the nature of areas is thus enriched. These attributes are particularly the change in the area or volume of deformable media, their density, and the local variation of density within these media.
REFERENCES:
patent: 5148500 (1992-09-01), Belanger
patent: 5173945 (1992-12-01), Pieters et al.
patent: 5224179 (1993-06-01), Denker et al.
patent: 5729465 (1998-03-01), Barbaresco
patent: 5734756 (1998-03-01), Sherman et al.
patent: 5757953 (1998-05-01), Jang
patent: 5966129 (1999-10-01), Matsukuma et al.
patent: 6058219 (2000-05-01), Patridge
patent: 6134353 (2000-10-01), Makram-Ebeid
patent: 6175655 (2001-01-01), George, III et al.
Hussein M. Yahia, et al., “Fast and Robust Level-Set Segmentation Of Deformable Structures,” Proceedings of the 1998 IEEE International Conference On Acoustics, Speech and Signal Processing, Seattle, WA., USA, 12-1, vol. 5, pp. 2765-2768.
S. Legoupil, et al., “Matching of Curvilinear Structures: Application to the Identification of Cortical Sulci on 3D Magnetic Resonance Brain Image,” Pattern Recognition in Practice IV: Multiple Paradigms, Comparative Studies and Hybrid Systems; Proceedings of an International Workshop; Vlieland, the Netherlands, Jun. 1-3, 1994; Elsevier, Amsterdam, pp. 185-195.
Brian c. S. Tom, et al., “Motion Estimation of Skeletonized Angiographic Images Using Elastic Registration,” IEEE Transactions On Medical Imaging, vol. 13, No. 3, (Sep. 1994), pp. 450-460.
P. Brigger, et al., “Morphological Shape Representation of Segmented Images Based on Temporally Modeled Motion Vectors, ” Proceedings on The International Conference On Image Processing, Austin, Nov. 13-16, 1994, vol. 3, No. Conf. 1, (Nov. 13, 1994), pp. 756-760, IEEE.
Ziheng Zhou, et al., “Generic Ribbons: A Morphological Approach Towards Natural Shape Decomposition,” Visual Communications and Image Processing IV, Philadelphia, PA, USA, Nov. 8-10, 1989, vol. 1199, pt. 1, pp. 170-180.
Barbaresco Frederic
Legoupil Samuel
Monnier Bernard
"Thomson-CSF"
Johns Andrew W.
Nakhjavan Shervin
LandOfFree
Process for dynamic monitoring of changes to deformable... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for dynamic monitoring of changes to deformable..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for dynamic monitoring of changes to deformable... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2962460