29 November 2023 in IEEE Signal Processing Letters
Recent years have seen great advances in the area of semantic segmentation. However, general methods are targeted at pinhole images and tend to underperform when directly adopted to panoramic images. And with the wide applications of panoramic cameras, it is important to develop feasible approaches to train segmentation models for their real-time applications. To address this problem, we propose a novel method using self-training and achieve comparable results on DensePASS dataset. Namely, we propose a deformable merge module tailored for panoramic images by efficiently and accurately incorporating features of different levels. We design a novel prototype adaptation term that aids the model to better learn the class-wise feature embeddings of distorted objects. Finally, we use a simple and valid evaluation method to achieve real-time and improved inference performance. All combined, we can reach 58.27% of mIoU scores on DensePASS dataset and achieve new state of the art results.
19 May 2025 in IEEE Transactions on Instrumentation and Measurement
This paper presents LONS-Calib, a novel extrinsic calibration framework for light detection and ranging (LiDAR) and omnidirectional cameras, specifically targeting panoramic annular lenses and fisheye lenses capable of capturing negative space information (z < 0). It also supports cameras with narrower fields of view, such as pinhole cameras. To overcome the impact of the significant distortion inherent in omnidirectional images, the framework optimizes chess-board corner extraction, projecting these corners onto a unit sphere for accurate camera pose estimation. Additionally, automatic detection of chessboard point clouds enhances efficiency in LiDAR pose estimation. The sample selection process is automated using the Variability of Quality (VOQ) score, which accounts for sensor measurement errors and the generalizability of calibration results. These features ensure that users, regardless of their calibration expertise, can easily record diverse poses and achieve high-quality calibration results with LONS-Calib. Extensive evaluations on synthetic and real-world datasets demonstrate remarkable calibration accuracy and efficiency, with typical optimization times under 0.3s on consumer-grade laptops. The calibration results also demonstrate applicability to multi-sensor fusion simultaneous localization and mapping (SLAM). Our framework is available at: https://github.com/wwendy233/LONS-Calib.