Limits...
Evaluation of a change detection methodology by means of binary thresholding algorithms and informational fusion processes.

Molina I, Martinez E, Arquero A, Pajares G, Sanchez J - Sensors (Basel) (2012)

Bottom Line: Those changes produce significant effects in human and natural activities.Then, the obtained results are evaluated by means of a quality control analysis, as well as with complementary graphical representations.The suggested methodology has also been proved efficiently for identifying the change detection index with the higher contribution.

View Article: PubMed Central - PubMed

Affiliation: ETSITGC, Universidad Politécnica de Madrid, Madrid, Spain. inigo.molina@upm.es

ABSTRACT
Landcover is subject to continuous changes on a wide variety of temporal and spatial scales. Those changes produce significant effects in human and natural activities. Maintaining an updated spatial database with the occurred changes allows a better monitoring of the Earth's resources and management of the environment. Change detection (CD) techniques using images from different sensors, such as satellite imagery, aerial photographs, etc., have proven to be suitable and secure data sources from which updated information can be extracted efficiently, so that changes can also be inventoried and monitored. In this paper, a multisource CD methodology for multiresolution datasets is applied. First, different change indices are processed, then different thresholding algorithms for change/no_change are applied to these indices in order to better estimate the statistical parameters of these categories, finally the indices are integrated into a change detection multisource fusion process, which allows generating a single CD result from several combination of indices. This methodology has been applied to datasets with different spectral and spatial resolution properties. Then, the obtained results are evaluated by means of a quality control analysis, as well as with complementary graphical representations. The suggested methodology has also been proved efficiently for identifying the change detection index with the higher contribution.

No MeSH data available.


Level 1 thresholded indices: best results (a) CVA–Li method (binary image); (b) NDVI–Otsu method (binary image).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3376593&req=5

f6-sensors-12-03528: Level 1 thresholded indices: best results (a) CVA–Li method (binary image); (b) NDVI–Otsu method (binary image).

Mentions: Always for level 1, in the case of the difference CD index based on NDVI’s, it is found that the information content is comprised between the value range [0,1], which implies that the change/no-change categories values are close to each other or within a short range of values. According to [14,47] the Otsu thresholding method provides satisfactory results when the measures of each category are similar, which is indeed the case for the classes contained in this index. These observations are formulated for the Ridler-Iter procedure [33] as well. While, the best results are achieved by these two methods, the overall accuracy does not reach in any case 80% and the individual accuracies for the change category are also not as expected (66.5–68.5%). Similarly, for change/no-change classes, producer’s (68.5% and 87.4%) and user’s (82.7% and 76.0%) accuracies also decrease, which may indicate that the NDVI CD index might not be suited for this particular geographical area. Nonetheless, the outcomes of these two procedures exhibit better a quality compared to those based on entropic algorithms. The result of the Otsu method is thus selected for further processes, as it shows a better overall accuracy and kappa coefficient. Figure 6 shows the best results obtained for level 1 in agreement with the quality values given previously. For the CVA case, it is confirmed that the radiometric differences tend to be overrated, while in the second case (NDVI difference index) these changes are not detected Figure 3(d). However, in both cases, the thresholded images have been solved adequately with the selected methods (Table 1).


Evaluation of a change detection methodology by means of binary thresholding algorithms and informational fusion processes.

Molina I, Martinez E, Arquero A, Pajares G, Sanchez J - Sensors (Basel) (2012)

Level 1 thresholded indices: best results (a) CVA–Li method (binary image); (b) NDVI–Otsu method (binary image).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3376593&req=5

f6-sensors-12-03528: Level 1 thresholded indices: best results (a) CVA–Li method (binary image); (b) NDVI–Otsu method (binary image).
Mentions: Always for level 1, in the case of the difference CD index based on NDVI’s, it is found that the information content is comprised between the value range [0,1], which implies that the change/no-change categories values are close to each other or within a short range of values. According to [14,47] the Otsu thresholding method provides satisfactory results when the measures of each category are similar, which is indeed the case for the classes contained in this index. These observations are formulated for the Ridler-Iter procedure [33] as well. While, the best results are achieved by these two methods, the overall accuracy does not reach in any case 80% and the individual accuracies for the change category are also not as expected (66.5–68.5%). Similarly, for change/no-change classes, producer’s (68.5% and 87.4%) and user’s (82.7% and 76.0%) accuracies also decrease, which may indicate that the NDVI CD index might not be suited for this particular geographical area. Nonetheless, the outcomes of these two procedures exhibit better a quality compared to those based on entropic algorithms. The result of the Otsu method is thus selected for further processes, as it shows a better overall accuracy and kappa coefficient. Figure 6 shows the best results obtained for level 1 in agreement with the quality values given previously. For the CVA case, it is confirmed that the radiometric differences tend to be overrated, while in the second case (NDVI difference index) these changes are not detected Figure 3(d). However, in both cases, the thresholded images have been solved adequately with the selected methods (Table 1).

Bottom Line: Those changes produce significant effects in human and natural activities.Then, the obtained results are evaluated by means of a quality control analysis, as well as with complementary graphical representations.The suggested methodology has also been proved efficiently for identifying the change detection index with the higher contribution.

View Article: PubMed Central - PubMed

Affiliation: ETSITGC, Universidad Politécnica de Madrid, Madrid, Spain. inigo.molina@upm.es

ABSTRACT
Landcover is subject to continuous changes on a wide variety of temporal and spatial scales. Those changes produce significant effects in human and natural activities. Maintaining an updated spatial database with the occurred changes allows a better monitoring of the Earth's resources and management of the environment. Change detection (CD) techniques using images from different sensors, such as satellite imagery, aerial photographs, etc., have proven to be suitable and secure data sources from which updated information can be extracted efficiently, so that changes can also be inventoried and monitored. In this paper, a multisource CD methodology for multiresolution datasets is applied. First, different change indices are processed, then different thresholding algorithms for change/no_change are applied to these indices in order to better estimate the statistical parameters of these categories, finally the indices are integrated into a change detection multisource fusion process, which allows generating a single CD result from several combination of indices. This methodology has been applied to datasets with different spectral and spatial resolution properties. Then, the obtained results are evaluated by means of a quality control analysis, as well as with complementary graphical representations. The suggested methodology has also been proved efficiently for identifying the change detection index with the higher contribution.

No MeSH data available.