Aerial Image Road Extraction Based on an Improved Generative Adversarial Network

Abstract

Aerial photographs and satellite images are one of the resources used for earth observation. In practice, automated detection of roads on aerial images is of significant values for the application such as car navigation, law enforcement, and fire services. In this paper, we present a novel road extraction method from aerial images based on an improved generative adversarial network, which is an end-to-end framework only requiring a few samples for training. Experimental results on the Massachusetts Roads Dataset show that the proposed method provides better performance than several state of the art techniques in terms of detection accuracy, recall, precision and F1-score.

Road Extraction of High-Resolution Remote Sensing Images Derived from DenseUNet

“…Su et al [33] enhanced the U-Net network model based on available problems. According to the characteristics of a small sample of aerial images, Zhang et al [34] proposed an improved network-based road extraction design framework. By refining the CNN architecture, Gao et al [35] proposed the refined deep residual convolutional neural network (RDRCNN) to enable it to detect the road area more accurately.…”

An Ensemble Wasserstein Generative Adversarial Network Method for Road Extraction From High Resolution Remote Sensing Images in Rural Areas

“…For example, [15] present a two-stage framework to extract roads, in which two GANs are used to detect roads and intersections followed by a smoothing-based optimization algorithm. [7] propose an improved GAN using the U-Net as G and suggest a simple loss function with an L2 loss and a cGAN loss. [18] create a multi-supervised GAN with two D to infer road networks with improved topology.…”

“…Each image x has an accompanying binary label y, indicating whether a pixel in the aerial image belongs to road (denoted as 1) or non-road (denoted as 0) as shown in Figure 4. To generate the label image, we follow the centerlinebased approach to label road vectors manually in ArcGIS software and then rasterize them with a line width of 5 pixels [7,39].…”

“…Most researchers regard road extraction from HR-RS images as a pixel-wise segmentation task which classifies each pixel into road or non-road according to geometric, photometric and texture features of roads (e.g. [2,7,8]). Thus, the performance of road extraction highly depends on the chosen semantic segmentation methods.…”

“…It consists of a generator G and a discriminator D where G generates images that are indistinguishable from the ground truth and D distinguishes between the ground truth and the generated images. Few works have used standard GANs in road extraction from HR-RS images and achieved good performances [4,7,15,17,18]. However, these standard GANs may suffer from vanishing gradients, mode collapse, and unstable training [19,20].…”