摘要
针对传统模型观测方法中存在的观测装置或传感器安装麻烦、工作量大、采样点有限等缺点。提出利用数字近景摄影测量技术量测模型变形。用高分辨率数码相机获取模型不同模拟开挖状态下的多组数字影像,采用改进的粒子群支持向量机方法自动检测和识别标志点,通过边缘跟踪和椭圆拟合算法确定标志点中心影像坐标。提出并推导了利用二维直接线性变换关系式和共线方程分解出影像外方位元素初值的实用算法,最后利用自检校光束法平差对标志点空间坐标进行精确定位。试验结果表明,空间坐标的测定精度高于±1 mm,所测位移场与实际变形情况比较一致,可以满足矿山岩层和地表移动相似材料模型观测要求。
The conventional model monitoring methods possess several shortcomings, such as complication of device or sensor installation, great workload and limited points of sampling etc. To overcome these shortcomings, the digital close-range photogrammetry technique was applied to monitor model deformation. The image groups of different simulated excavation status were captured by a high-resolution camera. An improved particle swarm optimization-based support vector machine （PSO-SVM） was used to detect and recognize the signalized points. The coordinates of the target centers were determined by contour tracking and ellipse fitting algorithm. Approximate values of the extrinsic parameters were decomposed using 2D direct linear transformation and collinearity equations. Bundle adjustment with self-calibration algorithm was used to determine the space coordinates of the targets. Test results indicate that the measurement accuracy is better than lmm. The displacement field measured by this technique is in good agreement with the real deformation, which can meet the requirements of model deformation monitoring of mine stratum and surface movements.
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2007年第6期1374-1379,共6页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金资助项目(50471046)
中国矿业大学青年科研基金资助项目(2005A030)
关键词
计算机应用
变形测量
相似材料模型
粒子群优化
支持向量机
自检校光束法平差
computer application
deformation measurement
simulated material model
particle swarm optimization
support vector machine
bundle adjustment with self-calibration
