互相关算法在运动目标距离选通激光三维成像中的应用

在微光成像中,距离选通成像是一种获取目标三维信息的有效手段。对于静止目标而言,可以通过对目标切片成像,获得不同选通距离的目标图像,进而通过二值化算法或质心算法获得目标的三维图像。而对于运动目标而言,在切片成像的同时,目标的空间位置会发生变化,因此需要对不同选通距离的目标图像进行配准,而后才能获得目标的三维图像。搭建了基于像增强电荷耦合器件(ICCD)的距离选通实验系统,提出了一种基于像质评价的互信息配准算法配准激光图像目标的空间位置,然后对配准后的激光图像通过互相关算法获得三维图像,并在保证一定的三维精度基础上,逐渐减少激光图像的像幅数,最后与二值化算法和质心算法三维重构进行了实验对比,实验结果表明,互信息配准算法能够有效配准激光图像,互相关三维成像算法精度高于二值化法和质心法,最少采用2幅图像即可获得目标的三维点云图像,大大降低了对运动目标三维成像的难度。

Application of cross-correlation algorithm in moving target range-gated three-dimensional laser active imaging

In low-light imaging, range-gated imaging is an effective method to obtain the 3D information of the target. As for the static targets, the different strobe distance target image can be obtained by the target slice imaging, and then the 3D image of the target is obtained by binary algorithm or centroid algorithm. For moving targets, in the same time of slice imaging, the spatial position of the target will change. Therefore, it is necessary to register different distance image before it can get a three-dimensional image of the object. Firstly,the range-gated ICCD experimental system was bulit, and mutual information image registration algorithm based on image quality evaluation was proposed for the image target registration. Then the 3D laser image was obtained by cross-correlation algorithm and the image number of laser image was gradually reduced on the basis of ensuring accuracy. Finally, the experiment of 3D reconstruction was carried out compared with the binary algorithm and centroid algorithm. Experimental results show that mutual information registration algorithm can effectively match laser image, mutual correlation algorithm for 3D imaging accuracy is higher than the binary method and centroid method. The 3D image can be obtained by using the two images, greatly reducing the difficulty of moving targets three-dimensional imaging.