Content area
Full Text
Abstract
Land cover/land use (LCLU) mapping is an important tool in automating site characterization for construction site preparation. By capturing multispectral images from an aerial perspective, LCLU mapping can be performed to characterize the construction site in a safe and efficient manner. The segmented LCLU maps can be used to determine the trafficability of the site and which construction equipment is best suited for clearing the site. While many traditional machine learning methods have been used for image segmentation, convolutional neural networks and deep learning approaches consistently outperform them. Three semantic segmentation models (PSPNet, U -Net, and Segnet) and three base models (VGG, ResNet, and MobileNet) are compared for the task of LCLU mapping. These models are pretrained on the ImageNet dataset and fine-tuned using datasets collected in central Illinois. The models are modified to include an additional channel for near-IR (NIR) images. Seven land cover classes (bare soil, water, roads/pavement, vegetation, trees, built-up, and an unknown category) were determined with an accuracy 82.71% by model ResNet/SegNet. Adding the NIR imagery achieved an accuracy of 74.74% by the semantic segmentation model VGG/PSPNet.
Keywords
Land Cover, Land Use, Deep Learning, Semantic Segmentation, Construction Site Analysis
1.Introduction
Land cover/land use (LCLU) mapping has traditionally been used in a variety of applications including flood forecasting, rangeland monitoring, climate change, biodiversity studies, resource management and land-use planning [1]. LCLU maps are an important tool for proper planning and use of natural resources and thus are applicable to construction site analysis and preparation [2]. Construction sites are notoriously hazardous and dirty and contain obstacles that may make it unsafe or impossible to characterize the site directly on foot [3]. The use of LCLU mapping in construction site characterization and preparation ensures the site can be characterized in a safe and efficient manner. The information provided by LCLU maps is vital in determining the trafficability of the site, and the major obstacles present. Using trafficability and obstacle information allows the user to designate the appropriate construction vehicles for site clearing and preparation.
The dramatic increase of free and open satellite imagery in the past few decades has enabled quick development of publicly available global coverage data [4]. Due to the evolution of remote sensing technologies and the recent increases in...