Comparison of CBF, ANN and SVM classifiers for object based classification of high resolution satellite images

Buddhiraju, Krishna Mohan; Rizvi, Imdad

doi:10.1109/igarss.2010.5652033

Cited by 25 publications

(12 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, these studies have been mostly conducted using pixel-based approaches. With the wide use of object-based approaches, there has been an increasing interest in comparing different machine learning classifiers using object-based methods [5,9,[15][16][17]. When using these machine learning classifiers, we should consider at least four key factors that can dramatically affect the classification accuracy and efficiency.…”

Section: Introductionmentioning

confidence: 99%

Comparing Machine Learning Classifiers for Object-Based Land Cover Classification Using Very High Resolution Imagery

et al. 2014

View full text Add to dashboard Cite

This study evaluates and compares the performance of four machine learning classifiers-support vector machine (SVM), normal Bayes (NB), classification and regression tree (CART) and K nearest neighbor (KNN)-to classify very high resolution images, using an object-based classification procedure. In particular, we investigated how tuning parameters affect the classification accuracy with different training sample sizes. We found that: (1) SVM and NB were superior to CART and KNN, and both could achieve high classification accuracy (>90%); (2) the setting of tuning parameters greatly affected classification accuracy, particularly for the most commonly-used SVM classifier; the optimal values of tuning parameters might vary slightly with the size of training samples; (3) the size of training sample also greatly affected the classification accuracy, when the size of training sample was less than 125. Increasing the size of training samples generally led to the increase of classification accuracies for all four classifiers. In addition, NB and KNN were more sensitive to the sample sizes. This research provides insights into the selection of classifiers and the size of training samples. It also highlights the importance of the appropriate setting of tuning parameters for different machine learning classifiers and provides useful information for optimizing these parameters.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparing Machine Learning Classifiers for Object-Based Land Cover Classification Using Very High Resolution Imagery

et al. 2014

View full text Add to dashboard Cite

show abstract

“…It can generate high accuracy for modeling complex nonlinear decision boundaries and is not easy to be over fitting. In fact, it is one of the most ideal algorithms for remote sensing classification [63].…”

Section: Support Vector Machinementioning

confidence: 99%

“…For better combination, the algorithms for base classifier training should be complementary to each other [57][58][59]. Support vector machine (SVM), C4.5 decision tree and artificial neural network (ANN) are among the most used remote sensing image classification algorithms, and are also quite different in land cover classification, thus have diversities [59][60][61][62][63]. For this reason, they are selected as the base classifiers here.…”

Section: Training Algorithms For Base Classifiersmentioning

confidence: 99%

Improving Land Use/Cover Classification with a Multiple Classifier System Using AdaBoost Integration Technique

2017

View full text Add to dashboard Cite

Guangzhou has experienced a rapid urbanization since 1978 when China initiated the economic reform, resulting in significant land use/cover changes (LUC). To produce a time series of accurate LUC dataset that can be used to study urbanization and its impacts, Landsat imagery was used to map LUC changes in Guangzhou from 1987 to 2015 at a three-year interval using a multiple classifier system (MCS). The system was based on a weighted vector to combine base classifiers of different classification algorithms, and was improved using the AdaBoost technique. The new classification method used support vector machines (SVM), C4.5 decision tree, and neural networks (ANN) as the training algorithms of the base classifiers, and produced higher overall classification accuracy (88.12%) and Kappa coefficient (0.87) than each base classifier did. The results of the experiment showed that, based on the accuracy improvement of each class, the overall accuracy was improved effectively, which combined advantages from each base classifier. The new method is of high robustness and low risk of overfitting, and is reliable and accurate, and could be used for analyzing urbanization processes and its impacts.

show abstract

“…Similarly the object-oriented methods produces better classifications because of its capacity of subdividing images into individual homogeneous regions (i.e., image objects) at scales that are appropriate to the inherent landscape (Rizvi and Buddhiraju, 2010) and establishing the context information and topological network of these image objects for accurate classification (Benz et al, 2004).…”

Section: Introductionmentioning

confidence: 99%

“…The classification is also better approach for shadow reconstruction, considering most spectroradiometric restoration algorithms were not designed to optimize classification performance, a support vector machine approach (Buddhiraju and Rizvi, 2010) was used to classify shadows for shadow pixels in satellite imagery (Panchal et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Evaluation of Shadow Reconstruction Algorithm for Very High Resolution Satellite Imagery

Jayant

Imdad

Kadam

2016

IJARSG

View full text Add to dashboard Cite

The Very High Resolution images have opened a new era for remote sensing applications such as object detection, classification, object mapping and change detection. However, the effects of shadow in these images are remarkable. The objective of this paper is to propose algorithm for shadow reconstruction. The primer stage, shadow detection and classification are carried out which separates shadow from rest of environment. For doing reconstruction the preprocessing is done by morphological operations which extract image components such as region and shape. It is followed by border creation and finally the reconstructed of an original image is accomplished border interpolation. The reconstruction accuracy is calculated by using Kappa Coefficient. Here Experimental results are obtained on three VHR images representing different shadow conditions. The performance analysis is carried out which shows better sensitivity and specificity.

show abstract

Comparison of CBF, ANN and SVM classifiers for object based classification of high resolution satellite images

Cited by 25 publications

References 6 publications

Comparing Machine Learning Classifiers for Object-Based Land Cover Classification Using Very High Resolution Imagery

Comparing Machine Learning Classifiers for Object-Based Land Cover Classification Using Very High Resolution Imagery

Improving Land Use/Cover Classification with a Multiple Classifier System Using AdaBoost Integration Technique

Evaluation of Shadow Reconstruction Algorithm for Very High Resolution Satellite Imagery

Contact Info

Product

Resources

About