摘要
研究空间运动目标到机动摄像机的距离的估计算法。通过相邻帧图像中同一目标上四对特征点的匹配,计算相邻采样时刻目标相对摄像机的距离变化量,结合目标的方位角、俯仰角、特征线度和摄像机的空间坐标,求解一个四阶非线性方程,得到前一时刻目标与摄像机之间的距离,随即确定出当前时刻上目标到摄像机距离。阐述了测距原理,推导了测距方程,最后给出了相应的实验数据。研究表明该方案通过两幅图像即可实现对目标的被动测距,缩比模型试验证明了算法的正确性和有效性。该方法适用于能够面成像的刚体三维运动目标。
Algorithm for estimating the distance from a moving target in space to the mobile camera was studied. The corresponding distance variation in the adjacent sampling time was calculated through matching 4-pairs points in adjacent frames in an image sequence. Then considering factors such as the target azimuth, pitching, characteristic linearity and the space coordinates of camera, the distance from an imaging target to the camera in former time could be obtained by solving a 4-order non-linearity equation on distance from an imaged target to the camera in former time. Consequently, the distance of the target from the camera at this time was obtained, too. The ranging principle was explored, the ranging equation was deduced, and experiment results were given along with the relative measurement data. Our research shows that this algorithm can estimate the distance from a moving target in space to the camera through only two images, and the experiments with scaled model are conducted to illustrate the effectiveness and the validity of the algorithm. This scheme adapts to any rigid target in three dimension movement.
出处
《光电工程》
EI
CAS
CSCD
北大核心
2007年第5期10-14,共5页
Opto-Electronic Engineering
关键词
距离测量
目标探测
图像匹配
光电成像
Distance measurement
Target detection
Image matching
Photoelectric imaging
