面向微尺度对象的多视场协同测量方法研究

摘要

针对显微视觉下视野小,无法对微小目标对象一次测量等难题,提出面向微尺度对象的多视场协同测量方法.该方法通过谱残差视觉注意模型建立亮度显著图对低分辨率的大视场图像提取兴趣区域,分别对兴趣区域获取高分辨率的小视场图像.然后通过ORB算法将大视场图像与小视场图像匹配起来,将小视场图像融合到大视场图像中.最后对得到的高分辨率图像进行图像处理,返回需要测量的数据.实验结果表明,该方法具有较高的测量精度,并且能够实现面向微尺度对象的自动测量.%Aiming at the problem that the micro visual horizons is small and cannot be measured at a small target all at once, a multiple filed synergy measurement method for micro scale object is proposed.This method extracts the region of interest from low resolution large FOV(field of view) image by spectral residual visual attention model establish brightness saliency map, and obtains the high resolution small FOV image from the region of interest.Then, the large FOV image is matched with the small FOV image by the ORB algorithm, and the small FOV image is fused into the large FOV image.Finally, the obtained high-resolution image is processed by image processing, and the data needed to be measured is returned.Experimental results show that the proposed scheme has high measurement accuracy, and can realize the automatic measurement for micro scale objects.