A Review of Landcover Classification with Very-High Resolution Remotely Sensed Optical Images—Analysis Unit, Model Scalability and Transferability

Abstract

As an important application in remote sensing, landcover classification remains one of the most challenging tasks in very-high-resolution (VHR) image analysis. As the rapidly increasing number of Deep Learning (DL) based landcover methods and training strategies are claimed to be the state-of-the-art, the already fragmented technical landscape of landcover mapping methods has been further complicated. Although there exists a plethora of literature review work attempting to guide researchers in making an informed choice of landcover mapping methods, the articles either focus on the review of applications in a specific area or revolve around general deep learning models, which lack a systematic view of the ever advancing landcover mapping methods. In addition, issues related to training samples and model transferability have become more critical than ever in an era dominated by data-driven approaches, but these issues were addressed to a lesser extent in previous review articles regarding remote sensing classification. Therefore, in this paper, we present a systematic overview of existing methods by starting from learning methods and varying basic analysis units for landcover mapping tasks, to challenges and solutions on three aspects of scalability and transferability with a remote sensing classification focus including (1) sparsity and imbalance of data; (2) domain gaps across different geographical regions; and (3) multi-source and multi-view fusion. We discuss in detail each of these categorical methods and draw concluding remarks in these developments and recommend potential directions for the continued endeavor.

A Review on Rainfall Measurement Based on Commercial Microwave Links in Wireless Cellular Networks

“…Bianchi et al [44] integrated the RG, radar and microwave links to improve the measurement accuracy of the spatial distribution of rainfall and rainfall intensity, and the Gauss-Newton method is used to minimize the cost function of all sensing methods. CML also has a great potential in the calibration of rainfall measured by radar or satellite images [106].…”

Evaluation of Different Satellite Images for Urban Land Use Analysis: A Case Study of Greater Manchester, United Kingdom

“…This is the process by which different classes or themes are extracted from raw remotely sensed digital satellite data [7,28,29]. Each pixel is usually considered to be a unique unit consisting of values in different spectral bands.…”

“…Each pixel is usually considered to be a unique unit consisting of values in different spectral bands. By comparing pixels to one another and to pixels of known identity, users of remotely sensed data can assemble groups of identical pixels into classes [7,30].…”

“…A land use land cover classification system like Anderson [20] was used for the classification. They are: (1) river (2) grassland (3) woodland (4) open space (5) motor way (6) railway (7) commercial buildings (8) residential buildings. It should be noted that the open class feature comprises of bare soils, construction sites and motor parks.…”

“…Remotely sensed data are usually deployed for image classification which can subsequently be used to describe the land cover of a region [3][4][5]. Although, remotely sensed images have been successfully used in mapping a wide range of land covers at varying spatial and temporal scales, their full potential as a land cover information source is yet to be realized [6,7].…”