Wen Yang鈥檚 Group Won the 2021 Best Paper Award of ISPRS J. P&RS Journal

        Recently, the work of line segment detection on arbitrary distorted images by the group of Professor Wen Yang in the Electronic Information School (EIS) and the group of Prof. Sebastian Scherer in the Robotics Institute at Carnegie Mellon University, won the 2021 Best Paper Award (U.V. Helava Award) of 銆奍SPRS Journal of Photogrammetry and Remote Sensing銆?

        The paper is titled 銆奤LSD: Unified line segment detection across pinhole, fisheye, and spherical cameras銆? Hao Li and Huai Yu are the co-first authors, Wen Yang and Lei Yu are both corresponding authors.

        This work proposed an end-to-end line detection network based on the Bezier curve representation, which can be directly applied to pinhole, fisheye and spherical images without image undistortion pre-processing. The Jury’s rationale for the  paper: This paper developed a supervised line detection method, valid either for distorted and undistorted images. The proposed model is therefore adapted to a large variety of sensors, as assessed by a comprehensive evaluation. The Jury acknowledges the significant effectiveness and generalizability of this work on a highly generic topic, coupled with the availability of both source code and dataset. Therefore, it very deserves the best paper award for 2021.

Award certificate

        Image line segment detection is a fundamental problem in computer vision and remote sensing. Although numerous state-of-the-art methods have shown great performance for straight line segment detection, line segment detection for distorted images without undistortion is still a challenging problem. Besides, there is a lack of a unified line segment detection framework for both distorted and undistorted images. To address these two problems, we propose a novel learning-based Unified Line Segment Detection method (i.e., ULSD) for distorted and undistorted images in this paper. Specifically, we first propose a novel equipartition point-based Bezier curve representation to model arbitrary distorted line segments. Then the line segment detection is tackled by equipartition point regression with an end-to-end trainable neural network. Consequently, the proposed ULSD is independent of camera distortion parameters and does not need any undistortion preprocessing. In the experiments, the proposed method is firstly evaluated on the pinhole, fisheye, and spherical image datasets, respectively, as well as trained and tested on the mixed dataset with differently distorted images. The experimental results on each distortion model show that the proposed ULSD is more competitive than the state-of-the-art methods for both accuracy and efficiency, especially for the results of the unified model trained on the mixed datasets, thus demonstrating the effectiveness and generality of the proposed ULSD to real-world scenarios. The source code is available at: https://github.com/lh9171338/ULSD-ISPRS

Pinhole, fisheye and spherical image line feature detection results

        The U.V. Helava Award, sponsored by Elsevier B.V. and Leica Geosystems AG, is a prestigious ISPRS Award, which was established in 1998 to encourage and stimulate submission of high quality scientific papers by individual authors or groups to the ISPRS Journal of Photogrammetry and Remote Sensing, to promote and advertise the Journal, and to honour the outstanding contributions of Dr. Uuno V. Helava to research and development in photogrammetry and remote sensing.