Multiscale and Multifeature Segmentation of High-Spatial Resolution Remote Sensing Images Using Superpixels with Mutual Optimal Strategy

Zhang, Fu; Sun, Yangjie; Fan, Liang; Han, Yutao

doi:10.3390/rs10081289

Cited by 28 publications

(21 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Over the past years, over 30 sorts of superpixel were developed to the public [60]. These algorithms can be generally divided into two categories: One is based on gradient ascent and the other one is based on graph theory [61]. Gradient ascent methods mainly cover the mean-shift (MS) algorithm [62], the simple linear iterative clustering (SLIC) algorithm [63] and the watershed transform algorithm [64]; while graph theory based methods mainly cover efficient graph-based image segmentation (EGB) [65] and the normalized cuts algorithm [66].…”

Section: Boundary Refinementmentioning

confidence: 99%

Fusion of Multiscale Convolutional Neural Networks for Building Extraction in Very High-Resolution Images

et al. 2019

View full text Add to dashboard Cite

Extracting buildings from very high resolution (VHR) images has attracted much attention but is still challenging due to their large varieties in appearance and scale. Convolutional neural networks (CNNs) have shown effective and superior performance in automatically learning high-level and discriminative features in extracting buildings. However, the fixed receptive fields make conventional CNNs insufficient to tolerate large scale changes. Multiscale CNN (MCNN) is a promising structure to meet this challenge. Unfortunately, the multiscale features extracted by MCNN are always stacked and fed into one classifier, which make it difficult to recognize objects with different scales. Besides, the repeated sub-sampling processes lead to a blurred boundary of the extracted features. In this study, we proposed a novel parallel support vector mechanism (SVM)-based fusion strategy to take full use of deep features at different scales as extracted by the MCNN structure. We firstly designed a MCNN structure with different sizes of input patches and kernels, to learn multiscale deep features. After that, features at different scales were individually fed into different support vector machine (SVM) classifiers to produce rule images for pre-classification. A decision fusion strategy is then applied on the pre-classification results based on another SVM classifier. Finally, superpixels are applied to refine the boundary of the fused results using region-based maximum voting. For performance evaluation, the well-known International Society for Photogrammetry and Remote Sensing (ISPRS) Potsdam dataset was used in comparison with several state-of-the-art algorithms. Experimental results have demonstrated the superior performance of the proposed methodology in extracting complex buildings in urban districts.

show abstract

Section: Boundary Refinementmentioning

confidence: 99%

Fusion of Multiscale Convolutional Neural Networks for Building Extraction in Very High-Resolution Images

et al. 2019

View full text Add to dashboard Cite

show abstract

“…In recent years, researchers have used many methods to extract road information from high-resolution remote sensing images. In this paper, referring to the method in references [50][51][52], the road area information of the 320 modeling samples was extracted from ZY3-02 high-resolution images and overlapped with the Luojia1-01 NTL imagery to reduce the interference of road light sources in the UBD estimation model. Figure 5 is a schematic diagram of the entire process.…”

Section: Ubd Estimation Model and Verificationmentioning

confidence: 99%

Correlation Analysis between UBD and LST in Hefei, China, Using Luojia1-01 Night-Time Light Imagery

et al. 2019

View full text Add to dashboard Cite

The urban heat island (UHI) is one of the essential phenomena of the modern urban climate. In recent years, urbanization in China has gradually accelerated, and the heat island effect has also intensified as the urban impervious surface area and the number of buildings is increasing. Urban building density (UBD) is one of the main factors affecting UHI, but there is little discussion on the relationship between the two. This paper takes Hefei as the research area, combines UBD data estimated by Luojia1-01 night-time light (NTL) imagery as the research object with land surface temperature (LST) data obtained from Landsat8 images, and carries out spatial correlation analysis on 0.5 × 0.5 km to 2 × 2 km resolution for them, so as to explore the relationship between UBD and UHI. The results show the following: (1) Luojia1-01 data have a good ability to estimate UBD and have fewer errors when compared with the actual UBD data; (2) At the four spatial scales, UBD and LST present a significant positive correlation that increases with the enlargement of the spatial scale; and (3) Moreover, the fitting effect of the Geographically Weighted Regression (GWR) model is better than that of the ordinary least squares (OLS) regression model. Appl. Sci. 2019, 9, 5224 2 of 20 absorbing heat faster and the specific heat capacity being smaller gives rise to limited heat release. In this way, the average LST in the area gradually increases with the spatial distribution of the UHI area to become more continuous and concentrated, forming a high-temperature area centered on artificial buildings in the city [1,9,10]. However, our comprehensive literature research found that past studies focused on the analysis of seasonal, daily, and annual variations of the UHI, which revealed the temporal and spatial variation of the UHI, the distribution characteristics of the horizontal and vertical directions, etc., while there has been less quantitative analysis of the single factors of UHIs, especially discussion of the ubiquitous relationship between impervious surfaces and UHI in the city.As an integral part of the urban impervious surface percentage, the urban building density (UBD) (percentage of built area) can directly reflect the degree of intensification and land-use efficiency in cities to some extent [11]. It is also an essential comprehensive social indicator that can function as an urban planning layout measure [12,13], resource utilization efficiency measure [14,15], and assessment of the urban ecological environment and livability [16][17][18][19]. Some researchers have pointed out that UBD is a crucial indicator to reduce the UHI effect [20] and proved a positive correlation between UBD and LST or UHI [21][22][23]. Additionally, researchers have tried to use a variety of urban design factors to study the relationships of UHIs, including UBD information. For example, Yang studied the summer heat island intensity of three high-rise residential districts in Shanghai, based on the three indices of building layout, density, and greeni...

show abstract

“…Therefore, image segmentation which is a necessary prerequisite step in GEOBIA and has a significant influence on the subsequent image processing should be able to solve the above problem. Recent years, researchers have focused on multiscale segmentation methods [10,43,44]. There is a strategy among them referred to as segment-merging which employs the edge information to produce initial partitions and then uses the interior information to conduct the merging stage and produce promising results.…”

Section: Discussionmentioning

confidence: 99%

Region Merging Method for Remote Sensing Spectral Image Aided by Inter-Segment and Boundary Homogeneities

et al. 2019

View full text Add to dashboard Cite

Image segmentation is extensively used in remote sensing spectral image processing. Most of the existing region merging methods assess the heterogeneity or homogeneity using global or pre-defined parameters, which lack the flexibility to further improve the goodness-of-fit. Recently, the local spectral angle (SA) threshold was used to produce promising segmentation results. However, this method falls short of considering the inherent relationship between adjacent segments. In order to overcome this limitation, an adaptive SA thresholds methods, which combines the inter-segment and boundary homogeneities of adjacent segment pairs by their respective weights to refine predetermined SA threshold, is employed in a hybrid segmentation framework to enhance the image segmentation accuracy. The proposed method can effectively improve the segmentation accuracy with different kinds of reference objects compared to the conventional segmentation approaches based on the global SA and local SA thresholds. The results of the visual comparison also reveal that our method can match more accurately with reference polygons of varied sizes and types.

show abstract

Multiscale and Multifeature Segmentation of High-Spatial Resolution Remote Sensing Images Using Superpixels with Mutual Optimal Strategy

Cited by 28 publications

References 49 publications

Fusion of Multiscale Convolutional Neural Networks for Building Extraction in Very High-Resolution Images

Fusion of Multiscale Convolutional Neural Networks for Building Extraction in Very High-Resolution Images

Correlation Analysis between UBD and LST in Hefei, China, Using Luojia1-01 Night-Time Light Imagery

Region Merging Method for Remote Sensing Spectral Image Aided by Inter-Segment and Boundary Homogeneities

Contact Info

Product

Resources

About