Accurate Instance Segmentation for Remote Sensing Images via Adaptive and Dynamic Feature Learning

Abstract

Instance segmentation for high-resolution remote sensing images (HRSIs) is a fundamental yet challenging task in earth observation, which aims at achieving instance-level location and pixel-level classification for instances of interest on the earth’s surface. The main difficulties come from the huge scale variation, arbitrary instance shapes, and numerous densely packed small objects in HRSIs. In this paper, we design an end-to-end multi-category instance segmentation network for HRSIs, where three new modules based on adaptive and dynamic feature learning are proposed to address the above issues. The cross-scale adaptive fusion (CSAF) module introduces a novel multi-scale feature fusion mechanism to enhance the capability of the model to detect and segment objects with noticeable size variation. To predict precise masks for the complex boundaries of remote sensing instances, we embed a context attention upsampling (CAU) kernel instead of deconvolution in the segmentation branch to aggregate contextual information for refined upsampling. Furthermore, we extend the general fixed positive and negative sample judgment threshold strategy into a dynamic sample selection (DSS) module to select more suitable positive and negative samples flexibly for densely packed instances. These three modules enable a better feature learning of the instance segmentation network. Extensive experiments are conducted on the iSAID and NWU VHR-10 instance segmentation datasets to validate the proposed method. Attributing to the three proposed modules, we have achieved 1.9% and 2.9% segmentation performance improvements on these two datasets compared with the baseline method and achieved the state-of-the-art performance.

Gaussian Dynamic Convolution for Semantic Segmentation in Remote Sensing Images

“…The framework consists of a double branch, partly used to suppress the large-scale background and partly used to activate the features of small objects. Some recent methods [5,27,28] try to incorporate modules that are effective in the field of general segmentation into remote sensing image segmentation networks, such as the well-known transformer or attention mechanisms, which are effective in improving the accuracy of the networks to some extent. However, these methods mainly target special application scenarios and are not effective in solving problems of semantic segmentation for high-resolution remote sensing images, such as multi-scale variation of objects and loss of foreground details.…”