Abstract
Random forest (RF) is a popular ensemble learning method that is widely used for the analysis of remote sensing images. RF also has connections with the kernel-based method. Its tree-based structure can generate an RF kernel (RFK) that provides an alternative to common kernels such as radial basis function (RBF) in kernel-based methods such as support vector machine (SVM). Using an RFK in an SVM has been shown to outperform both RF and SVM-RBF (i.e., using an RBF kernel in an SVM) in classification tasks with a high number of features. Here, we explore new designs of RFKs for remote sensing image classification. Different RF structural parameters and characteristics are used to generate various RFKs. In particular, we explore the use of RFs depth, the number of branches between terminal nodes of trees, and the predicted class probabilities for designing and evaluating new RFK. Two depth-based kernel are proposed: an RFK at the optimal depth, and a multiscale one created by combining RFKs at multiple depths. We evaluate the proposed kernels within an SVM by classifying a time series of Worldview-2 images, and by designing experiments having a various number of features. Benchmarking the new RFKs against the RBF shows that the newly proposed kernels outperform the RBF kernel for the experiments with a higher number of features. For the experiments with a lower number of features, RFKs and RBF kernel perform at about the same level. Benchmarking against the standard RF also shows the general outperformance of the proposed RFKs in an SVM. In all experiments, the best results are obtained with a depth-optimized RFK.
Highlights
R EMOTELY sensed images are one of the most important sources of data for land cover mapping
Comparing Random forest (RF) KBr and RF KNd The performance of the support vector machine (SVM)-RFKNd and SVM-RFKBr classifiers are compared in Table II that displays the overall accuracy (OA) versus
To overcome the downsides of classic RF kernel (RFK), we designed novel RFKs by using the distance among terminal nodes, obtaining classic RFK and RF-based class probabilities at multiple depths. We evaluated these novel kernels by comparing their performances in an SVM against classic radial basis function (RBF) and classic RFK for a crop classification problem over small-scale farms
Summary
R EMOTELY sensed images are one of the most important sources of data for land cover mapping. Producing high-quality land cover maps using remotely sensed data is still challenging because the necessary use of time series of images leads to high-dimensional problems and because land. Kernel-based methods map the nonlinear data into a reproducing kernel Hilbert space (RKHS) where the data is linearly separable. Instead of explicitly using a mapping function, a kernel function is used to reproduce the pairwise similarities matrix by computing the inner products among the samples in the RKHS [6]. Generative models aim at learning probability density functions and discriminative learning methods are based on learning class boundaries [19]. The key idea of discriminative kernels is that samples located in the same partition are similar and those ending up in different partitions are dissimilar [14]. We present the background on different possible RFKs designs selected for the experimental tests of this study
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.