Hyperspectral and Lidar Data Applied to the Urban Land Cover Machine Learning and Neural-Network-Based Classification: A Review

Abstract

Rapid technological advances in airborne hyperspectral and lidar systems paved the way for using machine learning algorithms to map urban environments. Both hyperspectral and lidar systems can discriminate among many significant urban structures and materials properties, which are not recognizable by applying conventional RGB cameras. In most recent years, the fusion of hyperspectral and lidar sensors has overcome challenges related to the limits of active and passive remote sensing systems, providing promising results in urban land cover classification. This paper presents principles and key features for airborne hyperspectral imaging, lidar, and the fusion of those, as well as applications of these for urban land cover classification. In addition, machine learning and deep learning classification algorithms suitable for classifying individual urban classes such as buildings, vegetation, and roads have been reviewed, focusing on extracted features critical for classification of urban surfaces, transferability, dimensionality, and computational expense.

An Object-Oriented Approach to the Classification of Roofing Materials Using Very High-Resolution Satellite Stereo-Pairs

“…The burdensome requirement of a large training set limits the adoption of deep learning algorithms for roofing materials classification, because in many practical cases ground-truth datasets are relatively small and they can be exploited more efficiently with SVM [51]. Here, the achieved accuracies (overall 91%, user and producer reported in Table 3) are comparable (or outperform in some cases) with the reported experiments using similar imagery and OBIA approach [12][13][14], and also with few works exploiting convolutional neural networks but with a more limited number of material classes [9,23].…”

Deep Learning in Multimodal Remote Sensing Data Fusion: A Comprehensive Review

“…Pansharpening (Ranchin et al, 2003) Introducing the methods belonging to ARSIS, along with giving a simple comparison (Vivone et al, 2014) Giving a thorough descripitions and assessments of the methods belonging to CS and MRA families (Meng et al, 2019) Introducing the methods belonging to CS, MAR, and VO from the idea of meta-analysis (Vivone et al, 2020) Giving a systematic introduction and evaluation of the methods in the category of CS, MAR, VO, and ML HS pansharpening (Loncan et al, 2015) Conducting a comprehensive analysis and evaluation in the methods from CS, MAR, hybrid, bayesian, and MF HS-MS fusion (Yokoya et al, 2017) Extensive experiments are presented to assess the methods from CS, MRA, unmixing, and bayesian (Dian et al, 2021b) Studying the performance of methods from CS, MAR, MF, TR, and DL Spatiotemporal (Chen et al, 2015) Discussing and evaluating four models from transformation/reconstruction/learning-based methods (Zhu et al, 2018) Reviewing the characteristics of five categories and their applications (Belgiu and Stein, 2019) Introducing the methods in three categories, as well as the challenges and opportunities (Li et al, 2020a) Analyzing the performance of representative methods with their provided benchmark dataset Heterogeneous fusion HS-LiDAR (Man et al, 2014) Summarizing the research on HS-LiDAR fusion for forest biomass estimation (Kuras et al, 2021) Giving an overview of HS-LiDAR fusion in the application of land cover classification SAR-optical (Kulkarni and Rege, 2020) Evaluating the performance of methods in CS and MRA in pixel-level RS-GBD (Li et al, 2021a) Providing a review on RS-social media fusion and their distributed strategies (Yin et al, 2021a) Reviewing the fusion of RS-GBD in the application of urban land use mapping from feature-level and decision-level perspectives Others (Wald, 1999) Setting up some definitions regrading data fusion (Gómez-Chova et al, 2015) Providing a review in seven data fusion applications for RS (Lahat et al, 2015) Summarizing the challenges in multimodal data fusion across various disciplines (Dalla Mura et al, 2015) Giving a comprehensive discussion on data fusion problems in RS by analyzing the Data Fusion Contests (Ghassemian, 2016) Introducing the RS fusion methods in pixel/feature/decision-level and different evaluation criteria (Schmitt and Zhu, 2016) Modeling the data fusion process, along with introducing some typical fusion scenarios in RS…”

A Two-Staged Feature Extraction Method Based on Total Variation for Hyperspectral Images

“…Therefore, hyperspectral images (HSIs) not only capture spatial features but also obtains rich spectral information from each pixel, which can achieve the classification and recognition of the target objects more efficiently than traditional images. Nowadays, many HSIs passively acquired on satellite or airborne have broad ranges of land cover; they are widely used in many fields such as urban mapping [1], agriculture [2], forest [3], and environmental monitoring [4]. In addition, HSIs can also be obtained by active remote sensing technology [5], which usually utilizes wide spectral light sources [6] to replace the sun to illuminate the scenes and which play a significant role in object detection [7] and recognition [8].…”