Image analysis – Applications – Motion or velocity measuring
Reexamination Certificate
2000-09-05
2004-09-28
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Applications
Motion or velocity measuring
C348S154000, C382S294000, C382S254000
Reexamination Certificate
active
06798897
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to apparatus and methods for registration, motion detection, tracking and background replacement.
BACKGROUND OF THE INVENTION
The state of the art is believed to be represented by the following publications:
H. S. Sawhney and R. Kumar. True Multi Image Alignment and its Application to Mosaicing and Lens Distortion. In Computer Vision and Pattern Recognition, pages 450-456, 1997;
P. Anandan. A Computational Framework and an Algorithm for the Measurement of Visual Motion. Int. J. of Computer Vision 2, pages 283-310, 1989;
M. Irani, B. Rousso and S. Peleg. Computing Occluding and Transparent Motions. Int. J. of Computer Vision, 12 No. 2, pages 5-16, January 1994;
E. Shilat, M. Werman and Y. Gdalyahu, Ridges' Corner Detection and Correspondence. In Computer Vision and Pattern Recognition, pages 976-981, 1997;
H. Wang and M. Brady. Real-Time Cornet Detection Algorithm for Motion Estimation. Image and Vision Computing 13 No. 9, pages 695-703, 1995;
Y. Rosenberg and M. Werman. Representing Local Motion as a Probability Matrix and Object Tracking. In Darpa Image Understanding Work Shop, pages 153-158, 1997;
M. Ben-Ezra, S. Peleg and M. Werman. Efficient Computation of the Most Probable Motion from Fuzzy Correspondences. Workshop on Application of Computer Vision, 1998;
M. Irani and P. Ananda, “Robust multi-sensor mage alignment”, Proceedings of International Conference on Computer Vision, January 1998.
“Blue-screen” background replacement is known.
U.S. Pat. No. 5,764,306 to Steffano describes a real time method of digitally altering a video data stream to remove portions of the original image and substitute elements to create a new image. Steffano describes real time replacement of the designated background portion of an incoming video signal with an alternate background. The actual background image is utilized for reference as the basis for determining the background and foreground elements within the image with the end result being comparable to traditional bluescreen processes, but requiring only a personal computer, video camera and software. The reference background image can be any reasonably static scene with a sufficient and stable light source captured by the camera. The video data stream is modified in real time by comparisons against the reference background image and is then passed on to its original destination. Multiple signal-noise processing algorithms are applied in real time against the signal to achieve a visually acceptable matte.
The disclosures of all publications mentioned in the specification and of the publications cited therein are hereby incorporated by reference.
SUMMARY OF THE INVENTION
The present invention seeks to provide a fast and robust method for image registration and motion detection based on discrete representation of the local motion. This allows the implementation of a real-time system on a PC computer which can register images and detect and track a moving object in video images, even when the camera is moving.
There is thus provided, in accordance with a preferred embodiment of the present invention, a method for registration between first and second images, the method including defining, for each individual location from among a plurality of locations sparsely distributed over the first image, a local probability matrix in which each element represents the probability of a possible displacement between he individual location in the first image and its corresponding location within the second image, defining a combined probability matrix by combining corresponding elements over the plurality of probability matrices, and computing an alignment of the first and second images in accordance with a combination of at least one of the largest of the elements of the combined probability matrix.
The above method is particularly suitable for applications in which the images can be assumed to be translated only relative to one another.
Further in accordance with a preferred embodiment of the present invention, the corresponding elements which are combined in the combined probability matrix defining step include elements within the local probability matrices which are similarly positioned if each individual local probability matrix is shifted to represent the effect on the individual location corresponding to the individual matrix, of a particular non-translational transformation between the first and second images.
The above method is particularly suitable for applications in which the images may be translated relative to one another and might be additionally rotated to a certain typically estimable extent.
Still further in accordance with a preferred embodiment of the present invention, the method also includes repeating the combined probability matrix defining step for each of a plurality of possible non-translational transformations between the first and second images, and selecting at least one most likely non-translational transformation from among the plurality of possible non-translational transformations, and the step of computing an alignment includes computing a relative non-translational transformation of the first and second images by computing a combination of the at least one most likely non-translational transformation, and computing a relative translation of the first and second images by computing a combination of at least one of the largest of the elements of the at least one combined probability matrices of the at least one most likely non-translational transformations.
Further in accordance with a preferred embodiment of the present invention, the step of selecting at least one most likely non-translational transformations from among the plurality of possible non-translational transformations includes comparing a set of at least one of the largest of the elements in each of the combined probability matrices of each of the plurality of possible non-translational transformations, selecting at least one set from among the compared sets whose members are largest, and selecting as most likely non-translational transformations, the at least one non-translational transformation corresponding to the at least one set whose members are largest.
Still further in accordance with a preferred embodiment of the present invention, the probability matrix is characterized in that each i,j element therewithin represents the probabilities that the individual point corresponds to an individual point in the second image, which is displaced correspondingly to the displacement of the i,j element from the center of the probability matrix.
Further in accordance with a preferred embodiment of the present invention, the corresponding elements which are combined include similarly positioned elements within the local probability matrices.
Still further in accordance with a preferred embodiment of the present invention, the method also includes executing the alignment.
Additionally in accordance with a preferred embodiment of the present invention, the method also includes executing the alignment by effecting the relative non-translational transformation and the relative translation of the first and second images.
Still further in accordance with a preferred embodiment of the present invention, the plurality of possible non-translational transformations between the first and second images includes at least one relative rotation between the first and second images.
Additionally in accordance with a preferred embodiment of the present invention, the plurality of possible non-translational transformations between the first and second images includes at least one relative zoom between the first and second images.
Further in accordance with a preferred embodiment of the present invention, the plurality of possible non-translational transformations between the first and second images includes at least one transformation which includes a combination of zoom and rotation between the first and second images.
Still further in accordance with a preferred embodiment of the present invention, the plurality of possible
Azarian Seyed
Mehta Bhavesh M.
Protrack Ltd.
LandOfFree
Real time image registration, motion detection and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Real time image registration, motion detection and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Real time image registration, motion detection and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3247820