条纹颜色分离与聚类  

Separating and clustering of structured light stripe color

作　　者：李华[1,2] 杨帆[1,2] 杨华民[1,2] 韩成[1,2] 张超[1,2] 

机构地区：[1]长春理工大学计算机科学技术学院,吉林长春130000 [2]特种电影技术及装备国家地方联合工程研究中心,吉林长春130000

出　　处：《光学精密工程》2016年第5期1206-1214,共9页Optics and Precision Engineering

基　　金：国家科技支撑计划资助项目(No.2012BAF12B22);吉林省重大科技攻关项目(No.2012ZDGG004);吉林省科技发展计划资助项目(No.20130303011GX;No.20140204050GX);长春市科技计划资助项目(No.14KG013)

摘　　要：为解决彩色物体表面颜色干扰以及条纹之间的颜色串扰导致彩色结构光系统产生的条纹颜色误识别问题,提出了彩色结构光的条纹颜色分离以及聚类方法。首先,提出了一种新的摄像机和投影机的颜色校正方法,以消除外界环境光以及系统的颜色串扰现象,解决系统的耦合性问题。其次,给出了一种物体的反射分离机制,用于提取物体表面的漫反射分量和镜面反射分量,修正彩色物体的背景色对彩色编码条纹的影响。最后,利用颜色聚类算法,对经过彩色物体调制后的条纹进行颜色归类划分,纠正条纹颜色之间的干扰,恢复系统消耗的彩色条纹颜色信息。实验验证了本文提出方法的有效性,结果表明:该方法具有较高的抗干扰能力、较强的鲁棒性,能准确有效地识别受干扰的彩色条纹,识别正确率达到92%左右。To overcome the misrecognition for the stripe color from a color structure light system caused by the interference of object surface color and the color crosstalk among the stripes,a method to separate and cluster the structured light stripe color was proposed.Firstly,a novel color correction method for a camera and a projector was designed to avoid the crosstalk phenomenon of color sequence from ambient light and the system and to solve the coupling issues of the system.Then,a reflective separation mechanism was presented to extract the diffuse component and the specular component of object surfaces and to correct the effect of object background color on color-coded stripes.Finally,the color clustering algorithm was used to classify the modulated stripe color and to correct theinterference among stripes and recover the information of stripe color due to energy consumption of the system.The feasibility of the proposed method was verified.Experimental results show that this method has higher anti-jamming capability and stronger robustness.It recognizes correctly the interfered colored stripes,and the recognition rate reaches about 92%.

关 键 词：计算机视觉 彩色结构光 彩色物体 反射分离 颜色分离 颜色聚类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN919.8[自动化与计算机技术—计算机科学与技术]