Remote sensing has become critical for understanding environmental dynamics, urban planning, and disaster management. However, traditional remote sensing workflows often rely on explicit segmentation or detection methods, which struggle to handle complex, implicit queries that require reasoning over spatial context, domain knowledge, and implicit user intent. Motivated by this, we introduce a new task, i.e., geospatial pixel reasoning, which allows implicit querying and reasoning and generates the mask of the target region. To advance this task, we construct and release the first large-scale benchmark dataset called EarthReason, which comprises 5,434 manually annotated image masks with over 30,000 implicit question-answer pairs. Moreover, we propose SegEarth-R1, a simple yet effective language-guided segmentation baseline that integrates a hierarchical visual encoder, a large language model (LLM) for instruction parsing, and a tailored mask generator for spatial correlation. The design of SegEarth-R1 incorporates domain-specific adaptations, including aggressive visual token compression to handle ultra-high-resolution remote sensing images, a description projection module to fuse language and multi-scale features, and a streamlined mask prediction pipeline that directly queries description embeddings. Extensive experiments demonstrate that SegEarth-R1 achieves state-of-the-art performance on both reasoning and referring segmentation tasks, significantly outperforming traditional and LLM-based segmentation methods.

Comparison between EarthReason and other related datasets. The gray rendering denotes the natural image dataset. "Seg", "Det", "VG", "Cls" denote segmentation, detection, visual grounding and classification datasets, respectively.

Overview of the proposed SegEarth-R1 architecture. Given an image and a text description, a hierarchical visual encoder and a proposed connector are used to extract and compress visual tokens. Then, the visual tokens and description embeddings are fed into an LLM for instruction interpretation and semantic correlation. Finally, description embeddings are directly mapped to the query vector and used for spatial correlation and segmentation mask generation.

Geospatial pixel reasoning results among SegEarth-R1 (ours) and previous related works.

Referring segmentation results among SegEarth-R1 and previous related works on RRSIS-D dataset.

Referring segmentation results among SegEarth-R1 and previous related works on RefSegRS dataset.

Qualitative Results of SegEarth-R1 on EarthReason.

Comparison with other models on EarthReason.

Comparison with PSALM on RRSIS-D.