Geocomplexity explains spatial errors

Zhang, Zehua; Song, Yongze; Luo, Peng; Wu, Peng

doi:10.1080/13658816.2023.2203212

Cited by 10 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For fine-scale remote sensing semantic segmentation, we defined complexity of geo-scenes as a statistical indicator of the context surrounding a target pixel or spatial location of interest. Our study focuses on understanding geocomplexity from the perspective of spatial variogram/dependence [34] and spatial heterogeneity [51]. Spatial variation or uncertainty can be intuitively measured by information entropy based on information theory, and its higher value indicates higher levels of uncertainty and randomness (and thus, a higher complexity), showing multiple classes located or clustered in a location of interest.…”

Section: Definition Of Geocomplexity Statistical Indicators and Compl...mentioning

confidence: 99%

“…Moran's I can be used to measure spatial dependence or distribution of spatial locations, and its value at a location reflects the spatial distribution pattern of its context or surrounding neighbors. Similar complexity measures based on information entropy and spatial statistics have also been proposed and led to important improvements in other remote sensing applications [32,34,57,58].…”

Section: Definition Of Geocomplexity Statistical Indicators and Compl...mentioning

confidence: 99%

“…The fractal dimension was employed in the ecological domain to understand spatial information [33]. Based on spatial neighbor dependence, a spatial local geocomplexity indicator was proposed to explain spatial errors between different spatial models [34]. It is a challenge to provide a geocomplexity indicator that can effectively improve the semantic segmentation of remote sensing.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Geocomplexity Statistical Indicator to Enhance Multiclass Semantic Segmentation of Remotely Sensed Data with Less Sampling Bias

He,

Li,

Gao

2024

Remote Sensing

View full text Add to dashboard Cite

Challenges in enhancing the multiclass segmentation of remotely sensed data include expensive and scarce labeled samples, complex geo-surface scenes, and resulting biases. The intricate nature of geographical surfaces, comprising varying elements and features, introduces significant complexity to the task of segmentation. The limited label data used to train segmentation models may exhibit biases due to imbalances or the inadequate representation of certain surface types or features. For applications like land use/cover monitoring, the assumption of evenly distributed simple random sampling may be not satisfied due to spatial stratified heterogeneity, introducing biases that can adversely impact the model’s ability to generalize effectively across diverse geographical areas. We introduced two statistical indicators to encode the complexity of geo-features under multiclass scenes and designed a corresponding optimal sampling scheme to select representative samples to reduce sampling bias during machine learning model training, especially that of deep learning models. The results of the complexity scores showed that the entropy-based and gray-based indicators effectively detected the complexity from geo-surface scenes: the entropy-based indicator was sensitive to the boundaries of different classes and the contours of geographical objects, while the Moran’s I indicator had a better performance in identifying the spatial structure information of geographical objects in remote sensing images. According to the complexity scores, the optimal sampling methods appropriately adapted the distribution of the training samples to the geo-context and enhanced their representativeness relative to the population. The single-score optimal sampling method achieved the highest improvement in DeepLab-V3 (increasing pixel accuracy by 0.3% and MIoU by 5.5%), and the multi-score optimal sampling method achieved the highest improvement in SegFormer (increasing ACC by 0.2% and MIoU by 2.4%). These findings carry significant implications for quantifying the complexity of geo-surface scenes and hence can enhance the semantic segmentation of high-resolution remote sensing images with less sampling bias.

show abstract

Section: Definition Of Geocomplexity Statistical Indicators and Compl...mentioning

confidence: 99%