单目图像中姿态角估计的坦克目标测距方法

徐超; 高敏; 曹欢

doi:doi:10.3788/gzxb20154405.0512002

光子学报, 2015, 44 (5): 0512002, 网络出版: 2015-05-26

单目图像中姿态角估计的坦克目标测距方法

The Attitude Angle Estimation-based Distance Measurement of Tank Target in Monocular Image

论文大纲

徐超 ^1,*高敏 ¹曹欢 ²

作者单位

¹ 军械工程学院导弹工程系, 河北石家庄 050003

² 北京航天嘉诚精密科技发展有限公司, 北京 102600

摘要

为了克服经典成像测距对相机姿态和视差信息依赖的问题, 以观瞄发控装置的电荷耦合器件获取的单目图像为研究对象, 提出了基于姿态角估计的坦克目标测距方法.该方法首先利用多级尺度不变特征变换匹配算法实现目标姿态角的快速估计, 获取与当前目标姿态偏差最小的模板; 然后利用该模板和目标图像模拟立体视觉, 建立不依赖于相机姿态和视差的单目测距模型.该模型在远距离测距和小角度偏差的前提下, 将模板中的匹配特征点投影到等效平面世界坐标系中, 建立被测坦克图像点及其空间点的映射关系, 将测距问题转化为对相机外参量的标定, 从而求解出相机中心到坦克中心的距离.采用200 m到2 500 m范围内的坦克验证测距方法的性能, 对测量误差随距离的变化规律进行分析, 并将其与不同射程时导弹末制导交班的最大容许误差进行对比.结果表明, 该方法适用于观瞄发控装置获取的任意姿态图像, 实现了野战条件下单兵反坦克**系统的远距离、快速测距; 测量结果中相对误差小于4.9%, 绝对误差小于87.7 m, 满足导弹各个射程的最大容许误差的要求.

Abstract

To overcome the traditional image ranging algorithm dependent on camera attitude and parallax, an attitude angle estimation-based ranging method of tank target, in monocular image cathched by observe control and launch system charge coupled device, is proposed. Firstly, multistage scale invariant feature transform matching method is applied to obtain the template with minor error comparing with current target by fast attitude angle estimation. Secondly, the template and current target are used to form a novel monocular range model, without camera attitude and parallax between two images. In such a model, the features points matching the tank target image of the template are projected to the approximated 2D world coordinate with the assumption of far distance and tiny error of the attitude. Thus we set up the mapping relationship between the image points and target points, and then external parameters of camera are achieved by monocular camera calibration to calculate the distance of the camera center from target center. Series of targets range from 200 m to 2 500 m are adopted to verify the effectiveness of the proposed algorithm. Sequentially, the alteration of error along with distance increasing is analyzed, and the comparison between measurement error and the maximal tolerance of middle-to-terminal guidance under various ranges is executed. The results indicates that the proposed algorithm do not need to capture the accurate position and angle of camera, resulting in fast and far distance measurement of individual anti-tank missile system in field. Meanwhile, the maximal relative error is 4.9% and the maximal absolute error is 87.7 m, which satisfy the demand of the biggest tolerance errors in the whole range of missile.

参考文献

[1] 王荻, 王洁, 付小宁. 基于单目图像和方向的测距系统及性能分析［J］. 光子学报, 2011, 40(7): 1112-1116.

WANG Di, WANG Jie, FU Xiao-ning. Ranging system and performance analysis based on monocular image and orientation［J］. Acta Photonica Sinica, 2011, 40(7): 1112-1116.

[2] 黄凤山, 刘恩福, 方忆湘, 等. 基于智能三坐标测量机的零件位姿单目立体视觉识别［J］. 光学·精密工程, 2013, 21(5): 1326-1332.

HUANG Feng-shan, LIU En-fu, FANG Yi-xiang, et al. Single camera stereo vision recognition for parts′ pose based on intelligent three coordinate measuring machine［J］. Optics and Precision Engineering, 2013, 21(5): 1326-1332.

[3] 姜广文, 晁志超, 伏思华, 等．基于可控旋转的像机高准确度标定技术［J］.光学学报, 2010, 30(5): 1308-1314.

JIANG Guang-wen, CHAO Zhi-chao, Fu Si-hua, et al. High- accurate camera calibration technique based on controllable rotation［J］. Acta Optical Sinica, 2010, 30(5): 1308-1314.

[4] 张法全, 路立平, 沈满德, 等. 单目视觉目标距离测量方法研究［J］. 光子学报, 2009, 38(2): 453-456.

ZHANG Fa-quan, LU LI-ping, SHEN Man-de, et al. Measurement method to object distances by monocular vision［J］. Acta Photonica Sinica, 2009, 38(2): 453-456.

[5] 王剑, 王元庆. 基于双焦成像的单眼立体视觉算法［J］. 机器人, 2007, 29(1): 41- 44.

WANG Jian, WANG Yuan-qing. A monocular stereo vision algorithm based on bifocal imaging［J］. Robot, 2007, 29(1): 41- 44.

[6] 张来线, 孙华燕, 郭惠超, 等. 基于图像灰度梯度最大值累加的自动调焦算法［J］. 光子学报, 2013, 42(5): 605-610.

ZHANG Lai-xian, SUN Hua-yan, GUO Hui-chao, et al. Auto focusing algorithm based on largest gray gradient summation［J］. Acta Photonica Sinica, 2013, 42(5): 605-610.

[7] 石春琴, 张丽艳, 韦虎, 等. 随机模式投影双目测量系统中的单目测量［J］. 光学·精密工程, 2010, 18(1): 257-265.

SHI Chun-qin, ZHANG Li-yan, WEI Hu, et al. Monocular measurement in binocular measurement system with instantaneous random illumination［J］. Optics and Precision Engineering, 2010, 18(1): 257-265.

[8] 吴刚, 唐振民. 单目式自主机器人视觉导航中的测距研究［J］. 机器人, 2010, 32(6): 828-832.

WU Gang, TANG Zhen-min. Distance measurement in visual navigation of monocular autonomous robots［J］. Robot, 2010, 32(6): 828-832.

[9] 霍炬, 崔家山, 王伟兴. 基于共面特征点的单目视觉位姿测量误差分析［J］. 光子学报, 2014, 43(5): 650-657.

HUO Ju, CUI Jia-shan, WANG Wei-xing. Error analysis of monocular visual position measurement based on coplanar feature points［J］. Acta Photonica Sinica, 2014, 43(5): 650-657.

[10] LI S, XU C, XIE M. A Robust O(n) Solution to the perspective-n-point problem［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7): 1444-1450.

[11] 韩延祥, 张志胜, 戴敏. 用于目标测距的单目视觉测量方法［J］. 光学精密工程, 2011, 19(5): 1110-1117.

HAN Yan-xiang, ZHANG Zhi-sheng, DAI Min. Monocular vision system for distance measurement based on feature points［J］. Optics and Precision Engineering, 2011, 19(5): 1110-1117.

[12] GONG J, FAN G, YU L, et al. Joint view-identity manifold for target tracking and recognition［C］. Image Processing (ICIP), 2012, 19th IEEE International Conference on IEEE, 2012: 1357-1360.

[13] GONG J, FAN G, YU L, et al. Joint target tracking, recognition and segmentation for infrared imagery using a shape manifold based level set［J］. Sensors, 2014, 14(6): 10124-10145.

[14] 熊英, 马惠敏. 3维物体SIFT特征的提取与应用［J］. 中国图象图形学报, 2010, 15(5): 814-819.

XIONG Ying, MA Hui-min. Extraction and application of 3D object SIFT feature［J］. Journal of Image and Graphics, 2010, 15(5): 814-819.

[15] LOWE D G. Distinctive image features from scale- invariant keypoints［J］. International Journal of Computer Vision, 2004, 60(2): 91-110.

[16] 淦文燕, 李德毅, 王建民.一种基于数据场的层次聚类方法［J］. 电子学报, 2006, 34(2): 258-262.

GAN Wen-yan, LI De-yi, WANG Jian-min. An hierarchical clustering method based on data fields［J］. Acta Electronica Sinica, 2006, 34(2): 258-262.

[17] ZHANG Z Y. A flexible new technique for camera calibration［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11): 1330-1334.

[18] 李达, 李少洪. 一种新的地心地固坐标系传感器配准算法［J］. 系统工程与电子技术, 2007, 29(1): 143-146.

LI Da, LI Shao-hong. New registration algorithm of sensor in each-centered earth-fixed (ECEF) coordinate system［J］. Systems Engineering and Electronics, 2007, 29(1): 143-146.

[19] TERUI F, KAMIMURA H, NISHIDA S. Motion estimation to a failed satellite on orbit using stereo vision and 3D model matching［C］. Proceedings of 9th International Conference on Control, Automation, Robotics and Vision, Singapore, USA: IEEE, 2006: 1-8.

徐超, 高敏, 曹欢. 单目图像中姿态角估计的坦克目标测距方法[J]. 光子学报, 2015, 44(5): 0512002. XU Chao, GAO Min, CAO Huan. The Attitude Angle Estimation-based Distance Measurement of Tank Target in Monocular Image[J]. ACTA PHOTONICA SINICA, 2015, 44(5): 0512002.

单目图像中姿态角估计的坦克目标测距方法

关于本站 Cookie 的使用提示

全站搜索

单目图像中姿态角估计的坦克目标测距方法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索