Multiscale Entropy-Based Surface Complexity Analysis for Land Cover Image Semantic Segmentation

Abstract

Recognizing and classifying natural or artificial geo-objects under complex geo-scenes using remotely sensed data remains a significant challenge due to the heterogeneity in their spatial distribution and sampling bias. In this study, we propose a deep learning method of surface complexity analysis based on multiscale entropy. This method can be used to reduce sampling bias and preserve entropy-based invariance in learning for the semantic segmentation of land use and land cover (LULC) images. Our quantitative models effectively identified and extracted local surface complexity scores, demonstrating their broad applicability. We tested our method using the Gaofen-2 image dataset in mainland China and accurately estimated multiscale complexity. A downstream evaluation revealed that our approach achieved similar or better performance compared to several representative state-of-the-art deep learning methods. This highlights the innovative and significant contribution of our entropy-based complexity analysis and its applicability in improving LULC semantic segmentations through optimal stratified sampling and constrained optimization, which can also potentially be used to enhance semantic segmentation under complex geo-scenes using other machine learning methods.

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

“…Land cover classes and the combination of different features were used as the stratified indicator to acquire high-quality training samples, exploring the impact of the training sample distribution on the accuracy of land cover classification [44]. Compared with our previous work [61], the multiclass scene typically exhibited a more complex geo-object distribution and structure compared to the binary class scene. A combination of geocomplexity indicators was used in our optimal sampling method.…”

“…The score of the target point or pixel reflected the entropy-based complexity of the surrounding context within the kernel of convolution. Thus, the kernel size was an important factor in complexity quantification [61]. ( )( ) ( )…”