A Robust Rule-Based Ensemble Framework Using Mean-Shift Segmentation for Hyperspectral Image Classification

Abstract

This paper assesses the performance of DoTRules-a dictionary of trusted rules-as a supervised rule-based ensemble framework based on the mean-shift segmentation for hyperspectral image classification. The proposed ensemble framework consists of multiple rule sets with rules constructed based on different class frequencies and sequences of occurrences. Shannon entropy was derived for assessing the uncertainty of every rule and the subsequent filtering of unreliable rules. DoTRules is not only a transparent approach for image classification but also a tool to map rule uncertainty, where rule uncertainty assessment can be applied as an estimate of classification accuracy prior to image classification. In this research, the proposed image classification framework is implemented using three world reference hyperspectral image datasets. We found that the overall accuracy of classification using the proposed ensemble framework was superior to state-of-the-art ensemble algorithms, as well as two non-ensemble algorithms, at multiple training sample sizes. We believe DoTRules can be applied more generally to the classification of discrete data such as hyperspectral satellite imagery products.Bayesian [17], maximum margin [18], evolutionary [19], clustering [9], association rule learning [20], artificial neural network [12,21,22] and deep learning [23] methods (see Figure 1). Regardless of the classification performance, many of these algorithms act as black-boxes, resulting in a poor recognition of the classification structure and robustness owing to the high-dimensionality of the data [24,25].

A Novel Method of Aircraft Detection under Complex Background Based on Circular Intensity Filter and Rotation Invariant Feature

“…Compared with other clustering methods, the mean-shift algorithm does not require the specification of the number of clusters. The mean-shift algorithm was first proposed by Fukunaga and Hostetler [ 23 ], and it is widely used in the fields of data clustering, image classification [ 24 ], image segmentation [ 25 ], target tracking [ 26 ], etc.…”

Real-Time Detection of Linear Structure Objects Using Mean Shift Segmentation and Heuristic Search

“…It is an efficient statistical iterative algorithm proposed by Fukunage [23]. And it is widely applied to image segmentation, object tracking, and image classification [24][25][26]. An image contains the linear structure object given in Figure 2(a) and the gradient orientation map is given in Figure 2(b), in which the gradient orientation value of every point is represented by gray level.…”

Image Segmentation and Object-Based Image Analysis for Environmental Monitoring: Recent Areas of Interest, Researchers’ Views on the Future Priorities

“…In December 2019, a Special Issue on the topic of GEOBIA was published in Remote Sensing, entitled "Image segmentation for environmental monitoring". The eight papers published in the special issue were largely representative of the current topics of interest within GEOBIA, covering image segmentation algorithm development [19,20] and segmentation parameter optimization strategies [21,22] as well as object-based image classification [23][24][25] and image fusion [26] methods.…”