A local approach to optimize the scale parameter in multiresolution segmentation for multispectral imagery

García, Fulgencio Cánovas; Alonso‐Sarría, Francisco

doi:10.1080/10106049.2015.1004131

Cited by 44 publications

(27 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous work with the image appears as Step 0 and is described in [26,27]. The aim of each step is:…”

Section: Methodsmentioning

confidence: 99%

“…The importance given to each layer, the weights given to the form and the importance given to the compactness and smoothness were not optimized, but remained invariant throughout the whole optimization process. This methodology and the results are fully explained in [27]. As a result of the segmentation, a set of 1,076,937 objects was obtained.…”

Section: Image Segmentationmentioning

confidence: 99%

See 1 more Smart Citation

Optimal Combination of Classification Algorithms and Feature Ranking Methods for Object-Based Classification of Submeter Resolution Z/I-Imaging DMC Imagery

García

Alonso‐Sarría

2015

Remote Sensing

View full text Add to dashboard Cite

Object-based image analysis allows several different features to be calculated for the resulting objects. However, a large number of features means longer computing times and might even result in a loss of classification accuracy. In this study, we use four feature ranking methods (maximum correlation, average correlation, Jeffries-Matusita distance and mean decrease in the Gini index) and five classification algorithms (linear discriminant analysis, naive Bayes, weighted k-nearest neighbors, support vector machines and random forest). The objective is to discover the optimal algorithm and feature subset to maximize accuracy when classifying a set of 1,076,937 objects, produced by the prior segmentation of a 0.45-m resolution multispectral image, with 356 features calculated on each object. The study area is both large (9070 ha) and diverse, which increases the possibility to generalize the results. The mean decrease in the Gini index was found to be the feature ranking method that provided highest accuracy for all of the classification algorithms. In addition, support vector machines and random forest obtained the highest accuracy in the classification, both using their default parameters. This is a useful result that could be taken into account in the processing of high-resolution images in large and diverse areas to obtain a land cover classification.Remote Sens. 2015, 7 4652

show abstract

“…Previous work with the image appears as Step 0 and is described in [26,27]. The aim of each step is:…”

Section: Methodsmentioning

confidence: 99%

Section: Image Segmentationmentioning

confidence: 99%

Optimal Combination of Classification Algorithms and Feature Ranking Methods for Object-Based Classification of Submeter Resolution Z/I-Imaging DMC Imagery

García

Alonso‐Sarría

2015

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The optimal similarity threshold was then automatically selected when the two curves first crossed, and in the case that curves crossed more than once, the lowest similarity threshold was selected. The objective was to achieve a trade-off between having homogeneous and distinct areas, i.e., low internal variance and autocorrelation values at the same time, as in previous studies [79][80][81].…”

Section: Retrieval Of Habitat Functional Typesmentioning

confidence: 99%

Biophysical Characterization of Protected Areas Globally through Optimized Image Segmentation and Classification

et al. 2016

View full text Add to dashboard Cite

Protected areas (PAs) need to be assessed systematically according to biodiversity values and threats in order to support decision-making processes. For this, PAs can be characterized according to their species, ecosystems and threats, but such information is often difficult to access and usually not comparable across regions. There are currently over 200,000 PAs in the world, and assessing these systematically according to their ecological values remains a huge challenge. However, linking remote sensing with ecological modelling can help to overcome some limitations of conservation studies, such as the sampling bias of biodiversity inventories. The aim of this paper is to introduce eHabitat+, a habitat modelling service supporting the European Commission's Digital Observatory for Protected Areas, and specifically to discuss a component that systematically stratifies PAs into different habitat functional types based on remote sensing data. eHabitat+ uses an optimized procedure of automatic image segmentation based on several environmental variables to identify the main biophysical gradients in each PA. This allows a systematic production of key indicators on PAs that can be compared globally. Results from a few case studies are illustrated to show the benefits and limitations of this open-source tool.

show abstract

“…The approach described above and variations thereof have been used in numerous studies in recent years [16][17][18][19][20][21][22][23][24][25][26][27][28]. Considerable efforts have been made to extend the original approach, e.g., to use multiple bands of input imagery [13,16], to consider optima at multiple scales [20,29,30], or to use GS for class specific optimization [26].…”

Section: Introductionmentioning

confidence: 99%

“…Considerable efforts have been made to extend the original approach, e.g., to use multiple bands of input imagery [13,16], to consider optima at multiple scales [20,29,30], or to use GS for class specific optimization [26].…”

Section: Introductionmentioning

confidence: 99%

On the Objectivity of the Objective Function—Problems with Unsupervised Segmentation Evaluation Based on Global Score and a Possible Remedy

2017

View full text Add to dashboard Cite

Abstract:Image segmentation is a crucial stage at the very beginning of many geographic object-based image analysis (GEOBIA) workflows. While segmentation quality is generally deemed of great importance, selecting adequate tuning parameters for a segmentation algorithm can be tedious and subjective. Procedures to automatically choose parameters of a segmentation algorithm are meant to make the process objective and reproducible. One of those approaches, and perhaps the most frequently used unsupervised parameter optimization method in the context of GEOBIA is called the objective function, also known as Global Score. Unfortunately, the method exhibits a hitherto widely neglected, yet severe source of instability, which makes quality rankings inconsistent. We demonstrate the issue in detail and propose a modification of the Global Score to mitigate the problem. This hopefully serves as a starting point to spark further development of the popular approach.

show abstract

A local approach to optimize the scale parameter in multiresolution segmentation for multispectral imagery

Cited by 44 publications

References 41 publications

Optimal Combination of Classification Algorithms and Feature Ranking Methods for Object-Based Classification of Submeter Resolution Z/I-Imaging DMC Imagery

Optimal Combination of Classification Algorithms and Feature Ranking Methods for Object-Based Classification of Submeter Resolution Z/I-Imaging DMC Imagery

Biophysical Characterization of Protected Areas Globally through Optimized Image Segmentation and Classification

On the Objectivity of the Objective Function—Problems with Unsupervised Segmentation Evaluation Based on Global Score and a Possible Remedy

Contact Info

Product

Resources

About