Skip to Main content Skip to Navigation
Theses

Fusion d'images en télédétection satellitaire

Abstract : Earth observation satellites provide multispectral and panchromatic data having different spatial, spectral, temporal, and radiometric resolutions. The fusion of a panchromatic image having high spatial but low spectral resolutions with multispectral images having low spatial but high spectral resolutions is a key issue in many remote sensing applications that require both high spatial and high spectral resolutions. The fused image may provide feature enhancement, and classification accuracy increase. These image processing techniques are known as pan-sharpening or resolution fusion techniques. In this thesis three algorithms are proposed for pansharpening. In the component substitution category, our main contribution consists in using the IHS transform and boosting the Green band in the vegetated areas. In this case, two algorithms were proposed. In the first one the vegetation is detected using the NDVI index and the boosting is done before the fusion process. In contrast, for the second algorithm the boosting is done after the fusion process and the vegetation is delineated using a new index (HRNDVI) proposed for high resolution images. HRNDVI is used in vegetation extraction even in the complex urban case where the vegetation is scattered. Hence, a new method, using HRNDVI, was proposed and tested to extract vegetation. The third pansharpening algorithm is included in the multiresolution category based on the NSCT transform. The improvement is assured by using a low number of decomposition levels for multispectral images and a high number of decomposition levels for the panchromatic image. This strategy allows getting satisfying visual and quantitative results. Moreover, the contribution of the thesis is also about the quality assessment of the obtained pansharpened images. Thus, a new protocol for evaluating the quality is proposed. According to the application under consideration, it can be adjusted to favour the spectral or the spatial qualities.
Complete list of metadatas

Cited literature [76 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00922646
Contributor : Kidiyo Kpalma <>
Submitted on : Sunday, December 29, 2013 - 1:00:12 PM
Last modification on : Thursday, January 7, 2021 - 4:33:11 PM
Long-term archiving on: : Saturday, March 29, 2014 - 10:10:13 PM

Identifiers

  • HAL Id : tel-00922646, version 1

Citation

Miloud Chikr El-Mezouar. Fusion d'images en télédétection satellitaire. Environmental Engineering. INSA de Rennes; Université Djillali Liabes de Sidi Bel Abbès, 2012. English. ⟨NNT : D12-31⟩. ⟨tel-00922646⟩

Share

Metrics

Record views

2483

Files downloads

8559