Analysis and Mapping of Natural Landscapes from Satellite Images
Using Morphological Filters

Méring, Catherine; Callot, Yann; Kemmouche, Akila

doi:10.1051/mmm:1996129

Cited by 8 publications

(5 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In such a model, the trees, due to their structure, are imaged with a great diversity, causing the high texture of the image of the elevation model ( Figure 4a). This image was processed using granulometric analysis (Mering, et al, 1996;Mering, Chopin, 2002;Kupidura et al, 2010;Kupidura, 2010Kupidura, , 2015, allowing the effective localization of areas with a high texture. The result of the analysis -a granulometric map (Figure 4b) -was binarized to obtain the mask of areas with a strong structure (Figure 4c): variant 4 ab m. In the following step, an alternate filtration by reconstruction (Serra, 1982;Soille, 1999) increased the classification accuracy by removing the single misclassified pixels in other classes of land -variant 4 ab m + f. This is a morphological filtering, which consists of alternating use of the opening and closing operations by the reconstitution.…”

Section: N Nmentioning

confidence: 99%

Testing of Land Cover Classification From Multispectral Airborne Laser Scanning Data

Bakuła¹,

Kupidura²,

Jełowicki³

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:Multispectral Airborne Laser Scanning provides a new opportunity for airborne data collection. It provides high-density topographic surveying and is also a useful tool for land cover mapping. Use of a minimum of three intensity images from a multiwavelength laser scanner and 3D information included in the digital surface model has the potential for land cover/use classification and a discussion about the application of this type of data in land cover/use mapping has recently begun. In the test study, three laser reflectance intensity images (orthogonalized point cloud) acquired in green, near-infrared and short-wave infrared bands, together with a digital surface model, were used in land cover/use classification where six classes were distinguished: water, sand and gravel, concrete and asphalt, low vegetation, trees and buildings. In the tested methods, different approaches for classification were applied: spectral (based only on laser reflectance intensity images), spectral with elevation data as additional input data, and spectro-textural, using morphological granulometry as a method of texture analysis of both types of data: spectral images and the digital surface model. The method of generating the intensity raster was also tested in the experiment. Reference data were created based on visual interpretation of ALS data and traditional optical aerial and satellite images. The results have shown that multispectral ALS data are unlike typical multispectral optical images, and they have a major potential for land cover/use classification. An overall accuracy of classification over 90% was achieved. The fusion of multi-wavelength laser intensity images and elevation data, with the additional use of textural information derived from granulometric analysis of images, helped to improve the accuracy of classification significantly. The method of interpolation for the intensity raster was not very helpful, and using intensity rasters with both first and last return numbers slightly improved the results.

show abstract

Section: N Nmentioning

confidence: 99%

Testing of Land Cover Classification From Multispectral Airborne Laser Scanning Data

Bakuła¹,

Kupidura²,

Jełowicki³

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…As the source of textural information, we propose a granulometric analysis of an image. Its efficiency has been researched and proven previously by several researchers (Mering et al, 1996;Mering and Chopin, 2002;Mering et al, 2010;Mering, 2011, Kupidura, 2010, 2015Kupidura et al, 2010, Szeszko, 2014. Granulometric analysis is based originally on a sequence of morphological openings (Haas et al, 1967) using structuring elements of sequentially increasing size.…”

Section: Fundamentals Of the Proposed Modelmentioning

confidence: 94%

Automatic Approach to VHR Satellite Image Classification

Kupidura¹,

Osińska-Skotak²,

Pluto-Kossakowska³

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Commission VII, WG VII/4 KEY WORDS: remote sensing, classification, texture analysis, granulometry, VHR, satellite image, vegetation indices ABSTRACT:In this paper, we present a proposition of a fully automatic classification of VHR satellite images. Unlike the most widespread approaches: supervised classification, which requires prior defining of class signatures, or unsupervised classification, which must be followed by an interpretation of its results, the proposed method requires no human intervention except for the setting of the initial parameters. The presented approach bases on both spectral and textural analysis of the image and consists of 3 steps. The first step, the analysis of spectral data, relies on NDVI values. Its purpose is to distinguish between basic classes, such as water, vegetation and non-vegetation, which all differ significantly spectrally, thus they can be easily extracted basing on spectral analysis. The second step relies on granulometric maps. These are the product of local granulometric analysis of an image and present information on the texture of each pixel neighbourhood, depending on the texture grain. The purpose of texture analysis is to distinguish between different classes, spectrally similar, but yet of different texture, e.g. bare soil from a built-up area, or low vegetation from a wooded area. Due to the use of granulometric analysis, based on mathematical morphology opening and closing, the results are resistant to the border effect (qualifying borders of objects in an image as spaces of high texture), which affect other methods of texture analysis like GLCM statistics or fractal analysis. Therefore, the effectiveness of the analysis is relatively high. Several indices based on values of different granulometric maps have been developed to simplify the extraction of classes of different texture. The third and final step of the process relies on a vegetation index, based on near infrared and blue bands. Its purpose is to correct partially misclassified pixels. All the indices used in the classification model developed relate to reflectance values, so the preliminary step of recalculation of pixel DNs to reflectance is required. Thanks to this, the proposed approach is in theory universal, and might be applied to different satellite system images of different acquisition dates. The test data consists of 3 Pleiades images captured on different dates. Research allowed to determine optimal indices values. Using the same parameters, we obtained a very good accuracy of extraction of 5 land cover/use classes: water, low vegetation, bare soil, wooded area and built-up area in all the test images (kappa from 87% to 96%). What constitutes important, even significant changes in parameter values, did not cause a significant declination of classification accuracy, which demonstrates how robust the proposed method is.

show abstract

“…Its efficiency has been researched and proven previously by several researchers (Mering et al, 1996;Mering and Chopin, 2002;Mering et al, 2010;Mering, 2011, Kupidura, 2010, 2015Kupidura et al, 2010, Szeszko, 2014. Granulometric analysis is based originally on a sequence of morphological openings (Haas et al, 1967) using structuring elements of sequentially increasing size.…”

Section: Fundamentals Of the Proposed Modelmentioning

confidence: 99%

Automatic Approach to VHR Satellite Image Classification

Kupidura

Osińska-Skotak

Pluto-Kossakowska

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

In this paper, we present a proposition of a fully automatic classification of VHR satellite images. Unlike the most widespread approaches: supervised classification, which requires prior defining of class signatures, or unsupervised classification, which must be followed by an interpretation of its results, the proposed method requires no human intervention except for the setting of the initial parameters. The presented approach bases on both spectral and textural analysis of the image and consists of 3 steps. The first step, the analysis of spectral data, relies on NDVI values. Its purpose is to distinguish between basic classes, such as water, vegetation and non-vegetation, which all differ significantly spectrally, thus they can be easily extracted basing on spectral analysis. The second step relies on granulometric maps. These are the product of local granulometric analysis of an image and present information on the texture of each pixel neighbourhood, depending on the texture grain. The purpose of texture analysis is to distinguish between different classes, spectrally similar, but yet of different texture, e.g. bare soil from a built-up area, or low vegetation from a wooded area. Due to the use of granulometric analysis, based on mathematical morphology opening and closing, the results are resistant to the border effect (qualifying borders of objects in an image as spaces of high texture), which affect other methods of texture analysis like GLCM statistics or fractal analysis. Therefore, the effectiveness of the analysis is relatively high. Several indices based on values of different granulometric maps have been developed to simplify the extraction of classes of different texture. The third and final step of the process relies on a vegetation index, based on near infrared and blue bands. Its purpose is to correct partially misclassified pixels. All the indices used in the classification model developed relate to reflectance values, so the preliminary step of recalculation of pixel DNs to reflectance is required. Thanks to this, the proposed approach is in theory universal, and might be applied to different satellite system images of different acquisition dates. The test data consists of 3 Pleiades images captured on different dates. Research allowed to determine optimal indices values. Using the same parameters, we obtained a very good accuracy of extraction of 5 land cover/use classes: water, low vegetation, bare soil, wooded area and built-up area in all the test images (<i>kappa</i> from 87% to 96%). What constitutes important, even significant changes in parameter values, did not cause a significant declination of classification accuracy, which demonstrates how robust the proposed method is.

show abstract

Analysis and Mapping of Natural Landscapes from Satellite Images Using Morphological Filters

Cited by 8 publications

References 3 publications

Testing of Land Cover Classification From Multispectral Airborne Laser Scanning Data

Testing of Land Cover Classification From Multispectral Airborne Laser Scanning Data

Automatic Approach to VHR Satellite Image Classification

Automatic Approach to VHR Satellite Image Classification

Contact Info

Product

Resources

About