基于激光扫描的犯罪现场图像还原
|
摘要
传统融合激光扫描和多视图影像的犯罪现场重建技术,使用相位式激光扫描仪测量犯罪现场数据,该扫描仪射程较短,适用小范围犯罪现场测量,导致犯罪现场图像还原结果不理想,提出基于激光扫描的犯罪现场图像还原方法。采用脉冲式激光扫描仪,利用相位偏移法扫描犯罪现场获取犯罪现场点云数据;去除点云数据中的异常点并进行点云配准,完成犯罪现场遮挡部分的重建;将犯罪现场点云数据视为离散二次曲面,计算激光扫描坐标与物体不同灰度点间的长度,利用激光扫描点云数据表达犯罪现场的远近景,实现犯罪现场图像还原。实验结果表明,所提方法的图像还原时间均值约为0. 8 min,图像还原的准确度高达99. 7%,可用于犯罪现场图像的高效、快速还原。
Traditionally,the technique of reconstructing crime scene is combined with laser scanning and multiview image methods. The phase laser scanner is used to measure the crime scene data. The scanner has a short range and is suitable for small-scale crime scene measurement. The result of restore crime scene image is unsatisfactory. A method of restore crime scene image based on laser scanning is proposed. The pulse laser scanner is used to scan the crime scene to obtain the point cloud data by phase shift method. The outliers in the point cloud data are removed and the point cloud registration is carried out to reconstruct the occlusion part of the crime scene. The point cloud data of the crime scene is regarded as discrete quadric surface,and the laser scanning coordinates are calculated to be different from the object. The length of gray points is used to express the long-range and near-range of the crime scene by laser scanning point cloud data,and the image restoration of the crime scene is realized. The experimental results show that the average time of image restoration is about 0. 8 min,and the accuracy of image restoration is 99. 7%. The proposed method can be used to restore crime scene images efficiently and quickly.
Traditionally,the technique of reconstructing crime scene is combined with laser scanning and multiview image methods. The phase laser scanner is used to measure the crime scene data. The scanner has a short range and is suitable for small-scale crime scene measurement. The result of restore crime scene image is unsatisfactory. A method of restore crime scene image based on laser scanning is proposed. The pulse laser scanner is used to scan the crime scene to obtain the point cloud data by phase shift method. The outliers in the point cloud data are removed and the point cloud registration is carried out to reconstruct the occlusion part of the crime scene. The point cloud data of the crime scene is regarded as discrete quadric surface,and the laser scanning coordinates are calculated to be different from the object. The length of gray points is used to express the long-range and near-range of the crime scene by laser scanning point cloud data,and the image restoration of the crime scene is realized. The experimental results show that the average time of image restoration is about 0. 8 min,and the accuracy of image restoration is 99. 7%. The proposed method can be used to restore crime scene images efficiently and quickly.
引文
[1]熊春宝,杨林,熊爱成,等.基于三维激光扫描异地控制法的深基坑变形监测[J].建筑技术,2016,47(9):777-780.
[2]罗寒,王建强,宋丹妮,等.融合激光扫描和多视图影像的模型重建技术[J].测绘科学,2017,42(4):123-128.
[3]薛婷,阮维鹏,张少杰.基于激光扫描的密集泡状流三维重建与优化[J].光学精密工程,2017,25(3):555-561.
[4]崔治,邓曙光,肖卫初.利用HSSIM和残差比阈值的3维激光扫描图像去噪[J].激光技术,2015,39(5):669-673.
[5]杨蒙蒙,万幼川,徐景中.激光扫描仪四棱塔镜误差分析及定位精度研究[J].中国激光,2015,42(9):316-323.
[6]丁小康,闫磊,孔建磊,等.基于二维激光与图像的人工林采育目标检测方法[J].林业科学,2015,51(7):129-135.
[7] MICHAEL C W,KING J A,HESTER R B.Confocal laser scanning microscopy and three-dimensional reconstruction of cell clusters in serous fluids.[J].Diagnostic Cytopathology,2015,17(4):272-279.
[8]张邦凤.果树快速定位采摘机器人设计-基于红外和激光扫描技术[J].农机化研究,2017,39(6):233-237.
[9]王华,娄小平,李伟仙,等.基于FPGA工具的关节臂激光扫描测头CMOS图像采集系统[J].工具技术,2017,51(6):112-116.
[10]褚宏亮,殷跃平,曹峰,等.大型崩滑灾害变形三维激光扫描监测技术研究[J].水文地质工程地质,2015,42(3):128-134.
[11]彭晨,余柏蒗,吴宾,等.基于移动激光扫描点云特征图像和SVM的建筑物立面半自动提取方法[J].地球信息科学学报,2016,18(7):878-885.
[12]郑力嘉,孙宇瑞,蔡祥.基于激光扫描3D图像的植物亏水体态辨识与萎蔫指数比较[J].农业工程学报,2015,31(2):79-86.
[13]李博阳,耿楠,张志毅.基于机器视觉的激光条纹扫描系统[J].计算机仿真,2015,32(6):241-246.
[14] WANG W,WU Z,ZENG H.Image distortion and its correction in linear galvanometric mirrors-based laser-scanning microscopy.[J]. Journal of Biomedical Optics,2015,20(5):56001.
[15]胡翠春,牛丽红,袁清钰,等.基于扫描相机的激光三维成像研究[J].光学技术,2017,43(3):248-251.
[16] GAO S J,SHENG L,WU Z Y,et al.Rapid restoration and real-time detection on spot image of atmospheric laser communication[J]. Optics&Precision Engineering,2015,23(8):2393-2399.
[17]夏元,薛春燕,吴艳,等.常见致病真菌所致角膜炎的激光扫描共焦显微镜图像特点分析[J].中华实验眼科杂志,2016,34(2):155-159.
[2]罗寒,王建强,宋丹妮,等.融合激光扫描和多视图影像的模型重建技术[J].测绘科学,2017,42(4):123-128.
[3]薛婷,阮维鹏,张少杰.基于激光扫描的密集泡状流三维重建与优化[J].光学精密工程,2017,25(3):555-561.
[4]崔治,邓曙光,肖卫初.利用HSSIM和残差比阈值的3维激光扫描图像去噪[J].激光技术,2015,39(5):669-673.
[5]杨蒙蒙,万幼川,徐景中.激光扫描仪四棱塔镜误差分析及定位精度研究[J].中国激光,2015,42(9):316-323.
[6]丁小康,闫磊,孔建磊,等.基于二维激光与图像的人工林采育目标检测方法[J].林业科学,2015,51(7):129-135.
[7] MICHAEL C W,KING J A,HESTER R B.Confocal laser scanning microscopy and three-dimensional reconstruction of cell clusters in serous fluids.[J].Diagnostic Cytopathology,2015,17(4):272-279.
[8]张邦凤.果树快速定位采摘机器人设计-基于红外和激光扫描技术[J].农机化研究,2017,39(6):233-237.
[9]王华,娄小平,李伟仙,等.基于FPGA工具的关节臂激光扫描测头CMOS图像采集系统[J].工具技术,2017,51(6):112-116.
[10]褚宏亮,殷跃平,曹峰,等.大型崩滑灾害变形三维激光扫描监测技术研究[J].水文地质工程地质,2015,42(3):128-134.
[11]彭晨,余柏蒗,吴宾,等.基于移动激光扫描点云特征图像和SVM的建筑物立面半自动提取方法[J].地球信息科学学报,2016,18(7):878-885.
[12]郑力嘉,孙宇瑞,蔡祥.基于激光扫描3D图像的植物亏水体态辨识与萎蔫指数比较[J].农业工程学报,2015,31(2):79-86.
[13]李博阳,耿楠,张志毅.基于机器视觉的激光条纹扫描系统[J].计算机仿真,2015,32(6):241-246.
[14] WANG W,WU Z,ZENG H.Image distortion and its correction in linear galvanometric mirrors-based laser-scanning microscopy.[J]. Journal of Biomedical Optics,2015,20(5):56001.
[15]胡翠春,牛丽红,袁清钰,等.基于扫描相机的激光三维成像研究[J].光学技术,2017,43(3):248-251.
[16] GAO S J,SHENG L,WU Z Y,et al.Rapid restoration and real-time detection on spot image of atmospheric laser communication[J]. Optics&Precision Engineering,2015,23(8):2393-2399.
[17]夏元,薛春燕,吴艳,等.常见致病真菌所致角膜炎的激光扫描共焦显微镜图像特点分析[J].中华实验眼科杂志,2016,34(2):155-159.