Lieven Verbeke | Ghent University (original) (raw)
Papers by Lieven Verbeke
International Journal of Remote Sensing, 2004
This paper focuses on a method to overcome some of the disadvantages that are related with the us... more This paper focuses on a method to overcome some of the disadvantages that are related with the use of artificial neural networks (ANNs) as supervised classifiers. The proposed method aims at speeding up network learning, improving classification accuracies and reducing variability on classification performance due to random weight initialization. This can be realized by transferring implicit knowledge from a previously
Journal of Applied Remote Sensing, 2011
ABSTRACT Stand density, expressed as the number of trees per unit area, is an important forest ma... more ABSTRACT Stand density, expressed as the number of trees per unit area, is an important forest management parameter. It is used by foresters to evaluate regeneration, to assess the effect of forest management measures, or as an indicator variable for other stand parameters like age, basal area, and volume. In this work, a new density estimation procedure is proposed based on wavelet analysis of very high resolution optical imagery. Wavelet coefficients are related to reference densities on a per segment basis, using an artificial neural network. The method was evaluated on artificial imagery and two very high resolution datasets covering forests in Heverlee, Belgium and Les Beaux de Provence, France. Whenever possible, the method was compared with the well-known local maximum filter. Results show good correspondence between predicted and true stand densities. The average absolute error and the correlation between predicted and true density was 149 trees/ha and 0.91 for the artificial dataset, 100 trees/ha and 0.85 for the Heverlee site, and 49 trees/ha and 0.78 for the Les Beaux de Provence site. The local maximum filter consistently yielded lower accuracies, as it is essentially a tree localization tool, rather than a density estimator.
Proximal Soil Sensing, 2010
Journal of Applied Remote Sensing, 2011
Soil Science Society of America Journal, 2009
... two EM38 orientations increased the information present in each orientation separately (Cockx... more ... two EM38 orientations increased the information present in each orientation separately (Cockx et al., 2007; Mankin and Karthikeyan, 2002) while ... Four validation indices were calculated: (i) the mean estimation error (MEE), (ii) the mean squared estimation error (MSEE), (iii) the ...
Remote Sensing of Environment, 2004
Remote Sensing of Environment, 2007
Remote Sensing Letters, 2011
International Journal of Remote Sensing, 2003
In remotely sensed images, mixed pixels will always be present. Soft classification defines the m... more In remotely sensed images, mixed pixels will always be present. Soft classification defines the membership degree of these pixels for the different land cover classes. Sub-pixel mapping is a technique designed to use the information contained in these mixed pixels to obtain a sharpened image. Pixels are divided into sub-pixels, representing the land cover class fractions. Genetic algorithms combined with
International Journal of Remote Sensing, 2006
International Journal of Remote Sensing, 2011
A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is int... more A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is introduced. The method rests on the idea that a neural network learning machine, trained on artificially generated input–target couples, can be used to efficiently process real SAR data. The explicit plus-point of the method is that it is trained with artificially generated data, reducing the demands put on real input data such as data quality, availability and cost price. The artificial data can be generated in such a way that they fit the particular ...
International Journal of Remote Sensing, 2004
The use of Artificial Neural Networks (ANNs) for the classification of remotely sensed imagery of... more The use of Artificial Neural Networks (ANNs) for the classification of remotely sensed imagery offers several advantages over more conventional methods. Yet their training still requires a set of pixels with known land cover. To increase ANN classification accuracy when few training data are available, an algorithm was applied that allows experience gained in previous classifications to be reused. The proposed method was evaluated by classifying a tropical savannah region in northern Togo using Landsat Thematic Mapper (TM) imagery. The presented approach reached a mean kappa coefficient that was significantly larger (at the 95% level) than that obtained after training networks with randomly initialized weights. Also, the observed variances on the obtained accuracies were significantly lower when compared to networks that were randomly initialized. Finally, Bhattacharyya (BH) distances were used to explain why some land cover classes benefit more from knowledge transfer than others.
International Journal of Remote Sensing, 2014
ABSTRACT This study tackles a common, yet underrated problem in remote-sensing image analysis: th... more ABSTRACT This study tackles a common, yet underrated problem in remote-sensing image analysis: the fact that human interpretation is highly variable among different operators. Despite current technological advancements, human perception and interpretation are still vital components of the map-making process. Consequently, human errors can considerably bias both mapping and modelling results. In our study, we present a web-based tool to quantify operator variability and to identify the human and external factors affecting this variability. Human operators were given a series of images and were asked to hand-digitize different point, line, and polygon objects. The quantification of performance variability was achieved using both thematic and positional accuracy measures. Subsequently, a series of questions related to demographics, experience, and personality were asked, and the answers were also quantified. Correlation and regression analysis was then used to explain the variability in operator performance. From our study, we conclude that: 1 humans were seldom perfect in visual interpretation; 2 some geographic objects were more complex to accurately digitize than others; 3 there was a high degree of variability among image interpreters when hand-digitizing the same objects; and 4 operator performance was mainly determined by demographic, non-cognitive, and cognitive personality factors, whereas external and technical factors influenced operator performance to a lesser extent. Finally, the results also indicated a gradual decline in performance over time, mimicking classical mental fatigue effects.
Geocarto International, 2008
Chinese Science Bulletin, 2007
International Journal of Remote Sensing, 2004
This paper focuses on a method to overcome some of the disadvantages that are related with the us... more This paper focuses on a method to overcome some of the disadvantages that are related with the use of artificial neural networks (ANNs) as supervised classifiers. The proposed method aims at speeding up network learning, improving classification accuracies and reducing variability on classification performance due to random weight initialization. This can be realized by transferring implicit knowledge from a previously
Journal of Applied Remote Sensing, 2011
ABSTRACT Stand density, expressed as the number of trees per unit area, is an important forest ma... more ABSTRACT Stand density, expressed as the number of trees per unit area, is an important forest management parameter. It is used by foresters to evaluate regeneration, to assess the effect of forest management measures, or as an indicator variable for other stand parameters like age, basal area, and volume. In this work, a new density estimation procedure is proposed based on wavelet analysis of very high resolution optical imagery. Wavelet coefficients are related to reference densities on a per segment basis, using an artificial neural network. The method was evaluated on artificial imagery and two very high resolution datasets covering forests in Heverlee, Belgium and Les Beaux de Provence, France. Whenever possible, the method was compared with the well-known local maximum filter. Results show good correspondence between predicted and true stand densities. The average absolute error and the correlation between predicted and true density was 149 trees/ha and 0.91 for the artificial dataset, 100 trees/ha and 0.85 for the Heverlee site, and 49 trees/ha and 0.78 for the Les Beaux de Provence site. The local maximum filter consistently yielded lower accuracies, as it is essentially a tree localization tool, rather than a density estimator.
Proximal Soil Sensing, 2010
Journal of Applied Remote Sensing, 2011
Soil Science Society of America Journal, 2009
... two EM38 orientations increased the information present in each orientation separately (Cockx... more ... two EM38 orientations increased the information present in each orientation separately (Cockx et al., 2007; Mankin and Karthikeyan, 2002) while ... Four validation indices were calculated: (i) the mean estimation error (MEE), (ii) the mean squared estimation error (MSEE), (iii) the ...
Remote Sensing of Environment, 2004
Remote Sensing of Environment, 2007
Remote Sensing Letters, 2011
International Journal of Remote Sensing, 2003
In remotely sensed images, mixed pixels will always be present. Soft classification defines the m... more In remotely sensed images, mixed pixels will always be present. Soft classification defines the membership degree of these pixels for the different land cover classes. Sub-pixel mapping is a technique designed to use the information contained in these mixed pixels to obtain a sharpened image. Pixels are divided into sub-pixels, representing the land cover class fractions. Genetic algorithms combined with
International Journal of Remote Sensing, 2006
International Journal of Remote Sensing, 2011
A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is int... more A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is introduced. The method rests on the idea that a neural network learning machine, trained on artificially generated input–target couples, can be used to efficiently process real SAR data. The explicit plus-point of the method is that it is trained with artificially generated data, reducing the demands put on real input data such as data quality, availability and cost price. The artificial data can be generated in such a way that they fit the particular ...
International Journal of Remote Sensing, 2004
The use of Artificial Neural Networks (ANNs) for the classification of remotely sensed imagery of... more The use of Artificial Neural Networks (ANNs) for the classification of remotely sensed imagery offers several advantages over more conventional methods. Yet their training still requires a set of pixels with known land cover. To increase ANN classification accuracy when few training data are available, an algorithm was applied that allows experience gained in previous classifications to be reused. The proposed method was evaluated by classifying a tropical savannah region in northern Togo using Landsat Thematic Mapper (TM) imagery. The presented approach reached a mean kappa coefficient that was significantly larger (at the 95% level) than that obtained after training networks with randomly initialized weights. Also, the observed variances on the obtained accuracies were significantly lower when compared to networks that were randomly initialized. Finally, Bhattacharyya (BH) distances were used to explain why some land cover classes benefit more from knowledge transfer than others.
International Journal of Remote Sensing, 2014
ABSTRACT This study tackles a common, yet underrated problem in remote-sensing image analysis: th... more ABSTRACT This study tackles a common, yet underrated problem in remote-sensing image analysis: the fact that human interpretation is highly variable among different operators. Despite current technological advancements, human perception and interpretation are still vital components of the map-making process. Consequently, human errors can considerably bias both mapping and modelling results. In our study, we present a web-based tool to quantify operator variability and to identify the human and external factors affecting this variability. Human operators were given a series of images and were asked to hand-digitize different point, line, and polygon objects. The quantification of performance variability was achieved using both thematic and positional accuracy measures. Subsequently, a series of questions related to demographics, experience, and personality were asked, and the answers were also quantified. Correlation and regression analysis was then used to explain the variability in operator performance. From our study, we conclude that: 1 humans were seldom perfect in visual interpretation; 2 some geographic objects were more complex to accurately digitize than others; 3 there was a high degree of variability among image interpreters when hand-digitizing the same objects; and 4 operator performance was mainly determined by demographic, non-cognitive, and cognitive personality factors, whereas external and technical factors influenced operator performance to a lesser extent. Finally, the results also indicated a gradual decline in performance over time, mimicking classical mental fatigue effects.
Geocarto International, 2008
Chinese Science Bulletin, 2007