Photogrammetric method of creating single image captured by sectional sensor through satellite acquisition

Technical cybernetics. Information technology. Computer facilities


Аuthors

Barabin G. V.1*, Gusev E. V.2**

1. Research proizvodstvennay firm «Infosystem», 16, 3rd Mytischinskaya Str., Moscow, 129626, Russia
2. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

*e-mail: gbarabin@gmail.com
**e-mail: kafedra610@yandex.ru

Abstract

The article describes the methods of building a single image from images bands obtained by separate CCD sensors satellite cameras with overlapping areas. The aim was to develop methods that provide the more accurate matching and focused on working with large images (of the order of several gigabytes) which have a small area of overlap. It is also necessary to obtain a refine model of satellite acquisition for new joined image. This refine model can be used for photomap building when necessary.
Proposed photogrammetric method taking into consideration the multivariate model of satellite acquisition is such a refine parametric model that all its images are matched in projection of Earth. The iterative Gauss-Newton method with some additions for matrix regularization was used to refine the model. This optimization problem works for minimize residuals in binding points in Earth projection. Support points on the Earth for stability of flying model refine was also added. For regularization task, two weight matrices for model parameters was built which are supporting and binding points using prior information about satellite parameters and points precision. Using backward photogrammetric formula single image in focal plane is obtained.
Offered photogrammetric method for image joining have more easily implemented algorithm. This is achieved by reducing the number of parameters to refine. The method also has good image joining precision and gives a good potential precision for image transformation in projection of the Earth. This is obtaining by using together binding and supporting points.
The methods can be applied to obtain a single image for automatic geometric correction from a set of neighboring images at the stage of the initial processing of satellite images. These methods belong to the level of processing 1B according to the international classification which means that it includes radiometric correction and geometric correction of systematic errors of CCD sensors in scanning system.
The advantage of this method is that it provides not only joint image but also new refine parameters for satellite acquisition model. These parameters provide good joining of images and coordinate positioning of joined image in Earth’s projection. So one can obtain more accurate photomap from source images.

Keywords:

satellite images geometric correction, image joining, photogrammetric method, satellite model parameters refine

References

  1. Zlobin V.K., Eremeyev V.V. Obrabotka ajerokosmicheskih izobrazhenij (Aerospace image processing), Moscow, Fizmatlit, 2006, 288 p.
  2. R. Gonzalez , R. Woods. Cifrovaja obrabotka izobrazhenij (Digital image processing), Moscow, Technosphere, 2005, 1072 p.
  3. Richard Szeliski. Computer vision: Algorithms and Applications. Springer-Verlag London Limited, 2011, 559 p.
  4. Gusev V.U. Vychislitel’naja matematika i kibernetika, Moscow, MAKS Press, 2013, pp. 102-104.
  5. Misganu Debella-Gilo, Andreas Kääb. Sub-pixel precision image matching for measuring surface displacements on mass movements using normalized cross-correlation, Remote Sensing of Enviroment, 2011, pp. 130-142.
  6. Gusev V.U. 11-aja Mezhdunarodnaja konferencija «Aviacija i kosmonavtika — 2012», St. Petersburg, Masterskaja pechati, 2012, pp. 253-254.
  7. Panteleev A.V., Letova T.A. Metody optimizacii v primerah i zadachah (Optimization methods in the examples and problems), Moscow, Vysshaja shkola, 2005, 544 p.
  8. Gomozov O.A. Metody i tehnologii geometricheskoj obrabotki kosmicheskoj videoinformacii ot optiko-jelektronnyh sistem vysokogo prostranstvennogo razreshenija (Methods and technologies of geometrical processing satellite images from optical-electronic systems with high spatial resolution), Phd, thesis, Ryazan, 2005, 153 p.

Download

mai.ru — informational site MAI

Copyright © 2000-2020 by MAI

Вход