Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2019, Vol. 16, No. 1, pp. 9-16
The problems of image superimposition with one-pixel accuracy
A.I. Alexanin
1 , M.A. Morozov
1 , E.V. Fomin
1 1 Institute of Automation and Control Processes FEB RAS, Vladivostok, Russia
Accepted: 21.12.2018
DOI: 10.21046/2070-7401-2019-16-1-9-16
The work is dedicated to the development of a new method of automatic superimposition of high spatial resolution images. Superimposition is based on a tight net of reference points on two images. A SURF algorithm is used to select the correct pairs of reference points for computing the displacement vectors. Displacement of each pixel of the image under correction is calculated by finding an average of the vectors after filtration in any pixel vicinity. It includes an estimation of computational accuracy. This method of operation is considered for two examples of image superimposition of the satellite “Resource-P”, Geoton-L1 unit. The superimposition possibility with one-pixel accuracy is demonstrated in the first example of forest images in a hilly area (the height variability is more 200 m). The image was not ortho-transformed and had a shooting angle of more than 30°. Superimposition of multisensory and panchromatic Geoton images received for the same scan job is considered in the second example, which demonstrates the image discrepancies. The discrepancies appeared due to chip splicing and a lack of coincidence of the axis of the sensors forming the multisensory images. A good geometrical coincidence of images could be demonstrated if the shooting angles of the sensors had been correctly taken into account.
Keywords: Resource-P, Geoton-L1, SURF, reference points, superimposition of images
Full textReferences:
- Aleksanin A. I., Krasnopeev S. M., Morozov M. A., Fomin E. V., Sovmeshchenie izobrazhenii s rossiiskikh sputnikov “Resurs-P” (Superimposition of images from Russian Resource-P satellite), Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2018, Vol. 15, No. 1, pp. 18–28.
- Kuznetsov A. E., Presnyakov O. A., Myatov G. N., Strukturnoe vosstanovlenie izobrazhenii ot mnogomatrichnykh skanernykh datchikov distantsionnogo zondirovaniya Zemli (Stitching of remote sensing images from staggered TDI CCD), Tsifrovaya obrabotka signalov, 2015, No. 3, pp. 29–36.
- Stroilov N. A., Nikitin A. V., Kurkina A. N., Bessonov R. V., Belinskaya E. V., Voronkov S. V., Metodiki nazemnoi vzaimnoi privyazki sistem koordinat zvezdnykh datchikov (Ground intercalibration methods of star sensor coordinate frames), Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2017, Vol. 14, No. 4, pp. 52–63.
- Bay H., Ess A., Tuytelaars T., Gool L. V., SURF: Speeded Up Robust Features, Computer Vision and Image Understanding, 2008, Vol. 110, No. 3, pp. 346–359.
- Harris C., Stephens M., A combined corner and edge detector, Proc. 4th Alvey Vision Conf., 1988, pp. 147–151.