CAJ | 학술논문

针对摄像机标定的棋盘格角点检测问题，分析了自动检测和人工检测的优缺点，结合Hessian矩阵理论和Harris算法，通过计算图像形状算子动态自适应地确定角点，再采用圆形对称模板自适应地剔除伪角点，最后对角点进行亚像素定位。通过几种流行的自适应检测算法和大量实验证明，该算法适用范围广，稳定性好，精度高，运行时间短，并且针对噪声影响鲁棒性好。
침대섭상궤표정적기반격각점검측문제，분석료자동검측화인공검측적우결점，결합Hessian구진이론화Harris산법，통과계산도상형상산자동태자괄응지학정각점，재채용원형대칭모판자괄응지척제위각점，최후대각점진행아상소정위。통과궤충류행적자괄응검측산법화대량실험증명，해산법괄용범위엄，은정성호，정도고，운행시간단，병차침대조성영향로봉성호。
To solve camera calibration checkerboard corner detection problem, the advantages and disadvantages about the automatic detection were analyzed, then Hessian matrix theory was combined with Harris algorithm, by calculateing the image shape, corners were dynamically determined, and by using circular symmetry template, false corners were adaptivly eliminated, at last the corners were sub-pixel located. The experimental results compared to several popular adaptive detection algorithms demonstrate that the algorithm is suitable to a wide range, and has good stability, high accuracy, short running time, and robustness against noise.