Feature Selection Solution with High Dimensionality and Low-Sample Size for Land Cover Classification in Object-Based Image Analysis

Abstract

Land cover information extraction through object-based image analysis (OBIA) has become an important trend in remote sensing, thanks to the increasing availability of high-resolution imagery. Segmented objects have a large number of features that cause high-dimension and low-sample size problems in the classification process. In this study, on the basis of a partial least squares generalized linear regression (PLSGLR), we propose a group corrected PLSGLR, known as G-PLSGLR, that aims to reduce the redundancy of object features for land cover identifications. Using Gaofen-2 images, the area of interest was segmented and sampled to generate small sample-size training datasets with 51 object features. The features selected by G-PLSGLR were compared against a guided regularized random forest (GRRF) in metrics of reduction rate, feature redundancy, and accuracy assessment of classification. Three indicators of overall accuracy (OA), user's accuracy (UA), and producer's accuracy (PA) were applied for accuracy assessment in this paper. The result shows that the G-PLSGLR achieved a reduction rate of 9.27 with a feature redundancy of 0.29, and a value of OA 90.63%. The GRRF achieved a reduction rate of 1.61 with a feature redundancy of 0.42, and a value of OA 85.56%. The PA of each land cover category was more than 95% using features selected by G-PLSGLR, while the PA ranged from 77 to 96% using features selected by GRRF. The UA of G-PLSGLR-selected features ranged from 70 to 80% except for grass land and bare land, which achieved 10% higher UA than GRRF-selected features. The G-PLSGLR method we proposed has the advantages of a large reduction rate, low feature redundancy, and high classification performance, which can be applied in OBIA-based land cover classification.

Separating Built-Up Areas from Bare Land in Mediterranean Cities Using Sentinel-2A Imagery

“…These findings indicate a feature selection problem that occurs regardless of the classification algorithm. An increased number of features, such as the image bands, generally improves the accuracy; however, it increases the amount of training samples required [24]. Considering the limited training sample availability in most conditions and the nonlinear response of the LCU classes across several bands, which is known as the "Hughes effect" [25], there is a need for a method to select a subset of relevant features from the original dataset to improve the classification process and achieve a dimension reduction [26].One alternative to overcome the above-mentioned problem is to perform linear transformation methods (such as principal components analysis (PCA) and independent component analysis (ICA)), or nonlinear algorithms (such as locality adaptive discriminant analysis (LADA) and multiple marginal fisher analysis (MMFA)), to remove the correlations and higher-order dependences in the image bands and use the produced components as input data for classification, to simplify and improve the process.…”

“…These findings indicate a feature selection problem that occurs regardless of the classification algorithm. An increased number of features, such as the image bands, generally improves the accuracy; however, it increases the amount of training samples required [24]. Considering the limited training sample availability in most conditions and the nonlinear response of the LCU classes across several bands, which is known as the "Hughes effect" [25], there is a need for a method to select a subset of relevant features from the original dataset to improve the classification process and achieve a dimension reduction [26].…”

A Deep Neural Networks Approach for Augmenting Samples of Land Cover Classification

“…Feature selection can improve the performance of object-based image classification [11]. To select features under low sample size and high dimensionality, we adapted a revised version of our previous work [14], which is designed to provide a solution to this problem. The previous work, namely the group-corrected partial least squares generalized linear regression (PLSGLR) method, can be described in three steps: (1) Group features based on Pearson's correlation coefficient; (2) rank features by PLSGLR and remove insignificant features; (3) reconstruct categories when the features are added one-by-one to calculate the Bayesian information criterion.…”

“…Compared with pixels that provides spectral information, image objects contain additional information on spectra, geometry, context, and texture. Thus, OBIA leads to sample sparsity in high dimensional data space, which increases the needs for larger sample size [14]. Although some machine learning algorithms, which are popular in supervised remote sensing classification, are tolerable to insufficient sample size in high dimensions, studies show that the sample size leads to larger variations in accuracy than the algorithms themselves [10,15].…”

Optimal Segmentation Scale Parameter, Feature Subset and Classification Algorithm for Geographic Object-Based Crop Recognition Using Multisource Satellite Imagery

“…Moreover, with the addition of texture and shape information, object-based classification can employ more features in recognizing target objects [10][11][12][13]. However, despite its advantages over traditional classifiers, the use of GEOBIA is also constrained by uncertainties associated with image segmentation [14], feature selection [15,16] and classification algorithms [17,18]. Segmentation is an essential process in GEOBIA, and the definition of segmentation parameters could have a significant influence on classification accuracy [19][20][21].…”