Road Extraction from VHR Remote-Sensing Imagery via Object Segmentation Constrained by Gabor Features

Chen, Li; Zhu, Qing; Xie, Xide; Han, Hu; Zeng, Haowei

doi:10.3390/ijgi7090362

Cited by 28 publications

(14 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With the development of DCNNs in recent years, many algorithms have been proposed for processing remote sensing images [25][26][27][28][29][30][31][32]. The fully convolutional network [33] (FCN) replaces the fully connected layers with convolutional layers, making it possible for large-scale dense prediction.…”

Section: Related Workmentioning

confidence: 99%

Joint Learning of Contour and Structure for Boundary-Preserved Building Extraction

Liao

Han

et al. 2021

Remote Sensing

Self Cite

View full text Add to dashboard Cite

Most of the existing approaches to the extraction of buildings from high-resolution orthoimages consider the problem as semantic segmentation, which extracts a pixel-wise mask for buildings and trains end-to-end with manually labeled building maps. However, as buildings are highly structured, such a strategy suffers several problems, such as blurred boundaries and the adhesion to close objects. To alleviate the above problems, we proposed a new strategy that also considers the contours of the buildings. Both the contours and structures of the buildings are jointly learned in the same network. The contours are learnable because the boundary of the mask labels of buildings implicitly represents the contours of buildings. We utilized the building contour information embedded in the labels to optimize the representation of building boundaries, then combined the contour information with multi-scale semantic features to enhance the robustness to image spatial resolution. The experimental results showed that the proposed method achieved 91.64%, 81.34%, and 74.51% intersection over union (IoU) on the WHU, Aerial, and Massachusetts building datasets, and outperformed the state-of-the-art (SOTA) methods. It significantly improved the accuracy of building boundaries, especially for the edges of adjacent buildings. The code is made publicly available.

show abstract

Section: Related Workmentioning

confidence: 99%

Joint Learning of Contour and Structure for Boundary-Preserved Building Extraction

Liao

Han

et al. 2021

Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recent studies show that the Gabor functions are gaining momentum in processing remote sensing images. Gabor filters were used to detect buildings ( [17] and land cover changes [18], and extract roads [19] from high resolution optical images. In [20], texture features from radar images were extracted with Gabor function and the output was used to identify rain and backscatter regions effectively.…”

Section: A Filtering With Gabor Functionmentioning

confidence: 99%

“…is Cartesian shear matrix Then, the scaled, sheared and translated trapezoidal wedges of corresponding cosine and sine Curvelet coefficients at each scale as reflection to each other and symmetry with respect to the origin are generated in the frequency domain (18). The Corresponding spatial domain is determined from (19).…”

Section: The Curvelet Transform Of the Mf-gabor Image Spacementioning

confidence: 99%

An Integrated Multiscale Geometric Analysis Approach for Automatic Extraction of Power Lines From High Resolution Remote Sensing Images

Feyissa

Cao

2020

IEEE Access

View full text Add to dashboard Cite

High resolution remote sensing systems provide cheaper and fast way of acquiring images of power lines. However, such images depicting the details of other complex background objects, noises, and complicated brightness measurements, make separate extraction of the power lines challenging. This paper addresses the problem of automatic extraction of power lines from high resolution remote sensing images obtained from different sources. In order to automatically extract the power lines, we proposed an integrated Multiscale Geometric Analysis (MGA) approach. First, complementary Gabor and matched filters (MF) were employed over an image to suppress unnecessary background and noises, and initial discrimination of the power lines. Then, the filtering output was decomposed in to scale and orientation based subband coefficients using the Fast Discrete Curvelet Transform (FDCT) so as to access and modify different image features separately. By employing selective modification operations, well-established power line structures ready for extraction were derived. Finally the powerlines were extracted with hysteresis thresholding. The approach was successful in extracting power lines from high resolution images captured in any orientation. It is robust even when the source image is cluttered, and degraded due to noise and brightness effects. Power lines represented by weak intensities, crossing bright image regions, changing direction, closer power lines and those crossing each other, disconnected/broken power lines due to noise and occlusions were all inferred and extracted successfully. The approach was validated using real test images and the performance measures showed over 90% average accuracy fitting the ground truth.

show abstract

“…Also, object-based classification is a popular classification method, where the element consists of a set of pixels (segments) rather than a single pixel. This method makes use of important characteristics (shape, texture, and contextual characteristics) that is solely in meaningful image objects and their common relationships [8]. The present work aims mainly to find the most suitable road network extraction approach that can be applied in the Egyptian environment.…”

Section: Introductionmentioning

confidence: 99%

Using of VHR Satellite Images for Road Network Extraction in Egypt

Nady

Mostafa

Abbas

et al. 2020

JES. Journal of Engineering Sciences

View full text Add to dashboard Cite

Roads extraction from VHR satellite images are very paramount for GIS and map updating. Due to the high resolution of satellite images, there are many obstacles broken roads such as shadow, and vehicles. The present work aims to find the most suitable road extraction approach that can be applied in the Egyptian environment. In this study, two satellite images from WorldView-2 and WorldView-3 were used. Classification of image by pixel-based and object-based was carried out to find the appropriate classification method for road extraction. Then, road class refinement by morphology and angular texture signature are performed to decrease the misclassifications between roads and other spectrally similar objects. After that, an iterative and localized Hough transform method was compared with the thinning algorithm method to find the proper method that can extract road centerline segments from the refined images. The performance of the extracted roads was estimated by using the common metrics; completeness, correctness, and quality. The results of this work demonstrate that the random tree in object-based classification achieves the highest overall accuracy than other classification methods. Also, thinning algorithm has more advantages than Hough transform.

show abstract

Road Extraction from VHR Remote-Sensing Imagery via Object Segmentation Constrained by Gabor Features

Cited by 28 publications

References 37 publications

Joint Learning of Contour and Structure for Boundary-Preserved Building Extraction

Joint Learning of Contour and Structure for Boundary-Preserved Building Extraction

An Integrated Multiscale Geometric Analysis Approach for Automatic Extraction of Power Lines From High Resolution Remote Sensing Images

Using of VHR Satellite Images for Road Network Extraction in Egypt

Contact Info

Product

Resources

About