Constructing Hierarchical Segmentation Tree for Feature Extraction and Land Cover Classification of High Resolution MS Imagery

Abstract

Accurate interpretation of high spatial resolution multispectral (MS) imagery relies on the extraction and fusion of information obtained from both spectral and spatial domains. Feature extraction from one or several fixed windows uses inaccurate description of pixel contexts and produces blurred object boundaries and low classification accuracy. In order to accurately characterize the spatial context properties of pixels, this paper presents a hierarchical-segmentation-based classification system. The system consists of two main modules: 1) hierarchical segmentation and 2) context-based classification. The segmentation module involves an optimization procedure to prevent undersegmentation of the land objects of interest and a scale selection procedure to find the most representative segmentation layers for modeling pixel contexts. The classification module couples a context-driven multilevel feature extraction methodology with a support vector machine classifier to get classification result. The proposed system is validated on three high spatial resolution MS data sets. Compared with state-of-the-art classification methods based on the similar concept, the proposed method demonstrates superior performance on both the classification accuracy and the quality of classification maps.

Exploring multiscale object-based convolutional neural network (multi-OCNN) for remote sensing image classification at high spatial resolution

“…The mean-shift algorithm presents the advantage of simple parametrization with great edge detection in high-resolution images (Wang et al, 2012;Ming et al, 2012;Su et al, 2015;Sun et al, 2019). Meanwhile, this segmentation tool has proven to be useful in both natural and urban contexts (Wang et al, 2015). The region growth is defined by specific homogeneity criterion and all pixels are grouping when they are closer in both spatial and spectral domain (Hossain and Chen, 2019).…”

Foreword to the Special Issue on Information Extraction From High-Spatial-Resolution Optical Remotely Sensed Imagery

“…In [16], an unsupervised local metric to quantify under-and oversegmentation is presented. Based on mean shift method, a hierarchical segmentation tree is constructed for context-based feature extraction [17].…”

“…As mentioned earlier, much work has been conducted from the perspective of feature extraction [18], [19], object-based image processing [17], etc. This section introduces five papers involving other important problems in high-resolution image classification.…”

Multitemporal Change Detection and Irregular Land Shape Area Measurement from Multispectral Sensor Images through BSO Algorithm

“…e SVM classifier classifies each pixel spatial information at different tree levels. e classifier extracts spatial context from image accurately [9]. e land cover type identifies trend component with phonological feature variability from time series of remote sensing images.…”