Illumination invariant change detection
View Patent ↗A system and method for illumination invariant change detection are provided, the system including a processor, an energy ranking unit in signal communication with the processor for extracting block coefficients for the first and second images and computing an energy difference responsive to the coefficients for a frequency energy between the first and second images, and a change detection unit in signal communication with the processor for analyzing the energy difference and detecting a scene change if the energy difference is indicative of change; and the method including receiving first and second images, extracting block coefficients corresponding to frequency energies for the first and second images, computing an energy difference for at least one of the frequency energies between the first and second images, analyzing the at least one energy difference, and detecting a scene change if the energy difference is indicative of change.
1. A method for illumination invariant change detection comprising:
using a computer to carry out the steps of: receiving first and second images;
extracting block coefficients corresponding to radial frequency energies for the first and second images;
scaling the radial frequency energies for one of the first or second images by a ratio of a first block coefficient of the other image to the first block coefficient of the one image;
computing at least one energy difference for at least one of the radial frequency energies between the first and second images;
analyzing the at least one energy difference; and
detecting a scene change if the energy difference is indicative of change.
2. A method as defined in claim 1 wherein:
computing at least one energy difference includes computing an energy scale responsive to the coefficients;
analyzing the at least one energy difference includes mapping a sum of the energy difference responsive to the energy scale; and
detecting a scene change is responsive to the frequency structure of the mapping.
3. A method as defined in claim 1 wherein the first and second images are processed in the feature space.
4. A method as defined in claim 3 wherein the feature space is one of the discrete cosine transform (DCT) domain and the wavelet domain.
5. A method as defined in claim 2 wherein the at least one energy difference is responsive to illumination change as a local contrast change.
6. A method as defined in claim 1 wherein the at least one energy difference is responsive to illumination change detection with a ranking approach.
7. A method as defined in claim 1 wherein the at least one energy difference is responsive to illumination change detection with a ranking approach and a sub-set ranking.
8. A method as defined in claim 1 wherein the at least one energy difference is responsive to detection with ranking only extremes DCT coefficients.
9. A method as defined in claim 4 wherein the feature space is the DCT domain and the images are compressed using at least one of JPEG, MPEG2, H.261, H.263, and MPEG4 compression standards.
10. A method as defined in claim 1 wherein the at least one energy difference is responsive to a multi-scale approach.
11. A method as defined in claim 3 wherein the images are received as non-compressed, further comprising performing a DCT transformation to the non-compressed image data for detection.
12. A method as defined in claim 1 wherein the at least one energy difference is responsive to a three-dimensional DCT space plus time approach.
13. A method as defined in claim 1 wherein the at least one energy difference is responsive to at least one of a gray level and color approach.
14. A system for illumination invariant change detection between first and second images comprising:
a processor;
an energy ranking unit in signal communication with the processor for extracting block coefficients for the first and second images and computing at least one energy difference responsive to the coefficients for at least one radial frequency energy between the first and second images, and scaling the radial frequency energies for one of the first or second images by a ratio of a first block coefficient of the other image to the first block coefficient of the one image; and
a change detection unit in signal communication with the processor for analyzing the at least one energy difference and detecting a scene change if the energy difference is indicative of change.
15. A system as defined in claim 14 , further comprising at least one of an imaging adapter and a communications adapter in signal communication with the processor for receiving image data.
16. A system as defined in claim 14 wherein the at least one energy difference is responsive to illumination change as a local contrast change.
17. A system as defined in claim 14 wherein the at least one energy difference is responsive to illumination change detection with a ranking approach.
18. A system as defined in claim 14 wherein the at least one energy difference is responsive to detection with ranking only extremes DCT coefficients.
19. A computer-readable medium tangibly embodying a computer program to perform program steps for illumination invariant change detection between first and second images, the program steps comprising:
extracting block coefficients corresponding to radial frequency energies for the first and second images;
scaling the radial frequency energies for one of the first or second images by a ratio of a first block coefficient of the other image to the first block coefficient of the one image;
computing at least one energy difference for at least one of the radial frequency energies between the first and second images;
analyzing the at least one energy difference; and
detecting a scene change if the energy difference is indicative of change.
20. A device as defined in claim 19 wherein the first and second images are processed in the feature space.