An Efficient Multi-Sensor Remote Sensing Image Clustering in Urban Areas via Boosted Convolutional Autoencoder (BCAE)

Maryam Rahimzad,Saeed Nadi,Saeid Homayouni,Amin Alizadeh Naeini

doi:10.3390/rs13132501

Abstract

High-resolution urban image clustering has remained a challenging task. This is mainly because its performance strongly depends on the discrimination power of features. Recently, several studies focused on unsupervised learning methods by autoencoders to learn and extract more efficient features for clustering purposes. This paper proposes a Boosted Convolutional AutoEncoder (BCAE) method based on feature learning for efficient urban image clustering. The proposed method was applied to multi-sensor remote-sensing images through a multistep workflow. The optical data were first preprocessed by applying a Minimum Noise Fraction (MNF) transformation. Then, these MNF features, in addition to the normalized Digital Surface Model (nDSM) and vegetation indexes such as Normalized Difference Vegetation Index (NDVI) and Excess Green (ExG(2)), were used as the inputs of the BCAE model. Next, our proposed convolutional autoencoder was trained to automatically encode upgraded features and boost the hand-crafted features for producing more clustering-friendly ones. Then, we employed the Mini Batch K-Means algorithm to cluster deep features. Finally, the comparative feature sets were manually designed in three modes to prove the efficiency of the proposed method in extracting compelling features. Experiments on three datasets show the efficiency of BCAE for feature learning. According to the experimental results, by applying the proposed method, the ultimate features become more suitable for clustering, and spatial correlation among the pixels in the feature learning process is also considered.

Highlights

Introduction published maps and institutional affilSatellite and airborne image classification is one of the most demanding remote sensing (RS) applications [1]
It should be noted that we used the normalized Digital Surface Model (nDSM) generated by Gerke [43]
It should be mentioned that Boosted Convolutional Autoencoder (BCAE) leads to compensate noise and extract the correct boundaries of the vast majority of classes

Summary

Introduction

Introduction published maps and institutional affilSatellite and airborne image classification is one of the most demanding remote sensing (RS) applications [1]. Image classification can be categorized as supervised and unsupervised approaches [2]. Supervised algorithms perform better than unsupervised ones, they require labeled or training samples. The unavailability of such high-quality and high-quantity training data justifies the use of clustering algorithms [3]. Another advantage of unsupervised methods is that they carry out the classification task by extracting valuable information from the data without any a priori knowledge or specific assumption about data. It is usually referred to as a subcategory of an unsupervised algorithm employed for dividing data into categories of a similar pattern without using training samples [6]. The efficiency of machine learning algorithms highly relies on the representation iations

Objectives

Methods

Results

Discussion

Conclusion