🤖 AI Summary
Arbitrary-oriented object detection in optical remote sensing imagery faces fundamental challenges including feature misalignment, spatial misalignment, and oriented bounding box (OBB) regression. This paper presents a systematic survey of deep learning advances from 2018 to 2023, proposing— for the first time—a three-dimensional taxonomy encompassing detection frameworks, OBB regression strategies, and feature representations. We introduce a unified analytical framework that clarifies key technical pathways: angular parameterization (e.g., sin/cos encoding, Gaussian modeling), rotation-aware RoI alignment, and remote sensing–specific evaluation protocols. Comprehensive benchmarking is conducted on DOTA and HRSC across CNN- and Transformer-based models, yielding a method capability map. The study provides both theoretical foundations and practical guidance for algorithm selection, novel model design, and standardization efforts in remote sensing object detection.
📝 Abstract
Oriented object detection is one of the most fundamental and challenging tasks in remote sensing, aiming to locate and classify objects with arbitrary orientations. Recent advancements in deep learning have significantly enhanced the capabilities of oriented object detection. Given the rapid development of this field, this paper presents a comprehensive survey of recent advances in oriented object detection. To be specific, we begin by tracing the technical evolution from horizontal object detection to oriented object detection and highlighting the specific challenges, including feature misalignment, spatial misalignment, and oriented bounding box (OBB) regression problems. Subsequently, we further categorize existing methods into detection framework, OBB regression, and feature representations, and provide an in-depth discussion on how these approaches address the above challenges. In addition, we cover several publicly available datasets and evaluation protocols. Furthermore, we provide a comprehensive comparison and analysis of state-of-the-art methods. Toward the end of this paper, we identify several future directions for oriented object detection.
