Abstract
Impervious surfaces are essential elements for the urban ecological environment. Machine-learning-based approaches have achieved successful breakthroughs in impervious surface extraction. These methods require large sets of labeled impervious surface data to train a model. However, it is a challenge to acquire massive impervious surface sample data because of complexity, time consumption, and high cost. To address this issue, we explore a method to generate massive impervious surface training samples using point of interest (POI) data and vehicle trajectory global positioning system data. Furthermore, a neural-network-based method is proposed for impervious surface extraction based on the generated training samples. One Landsat-8 image of Shenzhen City, China, was selected to test our approach. The extraction accuracy of the impervious surface was 90.88%, and the overall accuracy based on this method was improved by 8.57% and 8.45% compared with the support vector data description and weighted one-class support vector machine methods, respectively. The results show that the method integrating POI, trajectory data, and satellite imagery can be a viable candidate for impervious surface extraction.
Highlights
I MPERVIOUS surface constitutes the earth’s surface where water cannot penetrate, mainly including buildings made of various impermeable materials such as concrete and bricks
The weighted one-class support vector machine (WOC-SVM) assigning lower weights to the instances near the boundary of the data distribution to effectively weaken the effect of noise on the experiment
R,c,ζ vn ωiξi i s.t. ||φ(xi) − c||2 ≤ R2 + ξi ξi ≥ 0, i = 1, ..., n where ωi is the weight corresponding to the ith training sample, which is automatically generated from open data and contains both data frequency and spectral similarity features, calculated as in the equation (3)
Summary
I MPERVIOUS surface constitutes the earth’s surface where water cannot penetrate, mainly including buildings made of various impermeable materials such as concrete and bricks. The common methods for extracting impervious surfaces using remote sensing are spectral mixture analysis [6]–[9], the spectral index [10]–[12], and image classification method [13]–[15]. Spectral mixture analysis methods offer significant advantages in the reproducible and accurate classification of quantitative sub-pixel information These spectral mixture methods may produce an underestimation in the region of high-density impervious surfaces and overestimate the region of low-density impervious surfaces. The essence of extracting impervious surfaces using the spectral index involves the use of the differences in the spectral features to separate the effective features of impervious surfaces, which are applied to build the corresponding index models It is efficient and straightforward, but the method of determining classification thresholds requires further research. The study area is classified to determine whether the area is permeable or impervious, using common approaches such as neural network methods and support vector machine methods
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
