A camera calibration method combining FAST algorithm and double image beam adjustment

Journal Information

J. Hefei Univ. Tech. (Nat. Sci.), Jun.2025, Vol.48 No.6: 833-838

DOI: 10.3969/j.issn.1003-5060.2025.06.019

Authors

SONG Shengjia $ ^{1} $, ZHU Yongchao $ ^{1} $, SU Jie $ ^{2} $, LU Qiuling $ ^{1} $

(1. School of Civil and Hydraulic Engineering, Hefei University of Technology, Hefei 230009, China; 2. Lanzhou Tianrun Sifang Engineering Technology Co., Ltd., Lanzhou 730050, China)

Abstract and Keywords

Abstract: Non-measuring cameras are widely used in high-precision measurement because of their low price, small size and flexible use. However, camera calibration results have a great influence on the measurement accuracy, and the existing camera calibration methods have some problems such as insufficient accuracy or low efficiency. In this paper, a camera calibration method combining features from accelerated segment test (FAST) and double image beam adjustment is proposed. Firstly, the FAST algorithm was used to extract the marks automatically, and the initial values of the internal and external parameters of the camera were obtained by direct linear transformation method. Then, the relative parameters of the camera were calculated by the double image beam adjustment model based on the fixed baseline length constraint, and the convergence efficiency of the model was improved by the classification threshold method. The experimental results show that the proposed method has improved the accuracy of adjustment compared with the ordinary double image beam method, and the calibration error reaches 0.0064 mm. Moreover, the semi-automatic camera calibration process is realized, and the operation efficiency is improved. It is expected to be applied to the actual camera calibration operation.

Keywords: camera calibration; double image beam method; adjustment model; features from accelerated segment test(FAST) algorithm

作者信息

宋胜嘉 $ ^{1} $，朱勇超 $ ^{1} $，苏杰 $ ^{2} $，陆秋伶 $ ^{1} $

(1. 合肥工业大学土木与水利工程学院，安徽合肥 230009；2. 兰州天润四方工程技术有限公司，甘肃兰州 730050)

摘要和关键词

摘要: 非量测相机以价格低、体积小、使用灵活等优势被广泛应用于高精度测量工作中,但相机标定结果对测量精度影响较大,针对现有相机标定方法存在着精度不够或标定效率不高等问题,文章提出一种联合加速分割检测特征(features from accelerated segment test, FAST)算法和双像光束法平差的相机标定方法。首先利用FAST算法对像片标志点自动提取,利用直接线性变换方法解得相机内、外参数初值;然后基于固定基线长度约束的双像光束法平差模型解算相机相关参数,并通过分类阈值方法提高模型收敛效率。结果表明:该方法相较于普通双像光束法平差的精度有所提升,其标定后的中误差达到0.0064 mm,且实现了相机标定流程的半自动化,提高了作业效率,有望应用于实际场景的相机标定作业。

关键词: 相机标定；双像光束法；平差模型；FAST算法

Funding

国家自然科学基金资助项目（42104019；42171141）；中国科学院精密测量与技术创新研究大地测量与地球动力学国家重点实验室开放基金资助项目（SKLGED2022-3-7）

合肥工业大学学报(自然科学版)

Navigation