Multi-Scale Segmentation Method Based on Binary Merge Tree and Class Label Information

Su, Tengfei; Zhang, Shengwei

doi:10.1109/access.2018.2819988

Cited by 4 publications

(2 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To validate the proposed method, the scale parameters ranging from 6-25 were divided into three levels: the fine-scale level (6)(7)(8)(9)(10)(11)(12), medium-scale level (12)(13)(14)(15)(16)(17)(18), and coarse-scale level (18)(19)(20)(21)(22)(23)(24)(25). Then, 30 reference polygons were generated from each level at each test image to calculate the discrepancy metrics mentioned in Section Ⅱ.…”

Section: Comparative Analysis Of the Lp-based DV Methods 1) Multiplmentioning

confidence: 99%

“…Image segmentation is a key issue in GEOBIA since it can provide objects to GEOBIA by partitioning remote-sensing images into meaningful groups of pixels [9], and the objects are considered the minimum units of GEOBIA. Therefore, many scholars have paid attention to obtaining good objects for GEOBIA and have developed segmentation methods, such as the multiresolution segmentation [10], mean-shift segmentation [11], machine learning methods [12][13][14], and hybrid segmentation [8,[15][16][17][18][19]. However, defining appropriate segmentation parameters for good segmentation results is a major challenge [20,21], since the aforementioned algorithms almost entirely use certain user-defined parameters to control the segmentation quality [22].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Double-Variance Measures: A Potential Approach to Parameter Optimization of Remote Sensing Image Segmentation

Wang

Tian

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

The unsupervised segmentation evaluation (USE) method has been commonly used for remote-sensing segmentation parameter (SP) determinations to produce good segmentation results, due to its objectiveness and high efficiency. Existing studies have used different criteria to measure homogeneity and heterogeneity and have used certain combination strategies to form overall evaluations. However, different criteria have unique statistical characteristics. The differentiated statistical characteristics maintained in homogeneity and heterogeneity calculations may result in inherent instability in the USE results, leading to unsuitable SP selections. Moreover, few studies have focused on the simultaneous determination of a single optimal SP and multiple optimal SPs. In this paper, double-variance (DV) measures were proposed for recognizing more suitable SPs. Then, two combination strategies, F-measure and local peak (LP), were applied to test the potential of using double-variance measures to determine a single SP and multiple SPs, respectively. The multiresolution segmentation (MRS) algorithm and Gaofen-1 data were used to test the proposed method. The comparative results indicated that the DV is a more promising internal homogeneity and external heterogeneity metric for segmentation evaluation and optimal SP determination compared to conventional methods. The F-measure-based DV method could produce better overall goodness of segmentation for differently sized natural geo-objects, compared with the competing methods. The LP-based DV method could obtain multiple optimal scales that produced better segments for the identification of small, natural geo-objects to large, semantic geo-objects, compared to the competitive methods.

show abstract