一种基于图像融合的多线结构光立体视觉测量方法

A Measurement Method of Multi-line Structured Light Stereo Vision Based on Image Fusion

针对现有线结构光立体视觉测量系统在实际应用中,由于图像遮挡、反射等原因造成的缺陷导致精度降低和测量错误,提出一种基于多摄像机的图像融合的测量方法。在现有的多线结构光立体视觉测量系统的硬件平台上,经由提取光带图像,图像特征匹配,以立体视觉测量原理匹配光带图像和产生光带图像的线结构光刀,通过坐标转换和图像的重新采样插值将光带图像转换到同一个摄像机的成像平面进行图像融合,经试验得到的图像能够大幅降低由于图像缺陷导致的精度降低和测量错误。

A measurement method of image fusion based on multi camera was proposed to avoid reduced accuracy and measurement error caused by mage occlusion,reflection and other reasons in multi-line structured light stereovision measurement system. The measurement method can be used on the prior multi-line structured light stereo vision measurement system hardware platform by updating its software. The fusion image can be created after these processes: extracting band images,image feature matching,matching band image with the line structure light knife which create the band image by using traditional stereo vision measurement method,converting all band images to a camera's imaging plane,and realizing image fusion by resampling and interpolation of band images after coordinate transformation. Experiment proved that the method can greatly decrease reduced accuracy and measurement error caused by band image defects.