Department of Physics, Gottwald Science Center, University of Richmond, VA 23173, USA.
Ultramicroscopy. 2010 Mar;110(4):339-49. doi: 10.1016/j.ultramic.2010.01.006. Epub 2010 Jan 20.
A common source of distortion in scanning probe microscope (SPM) images is "thermal drift," the slow thermal expansion of different materials in the sample and microscope due to small changes in temperature over the course of a scan. We describe here a method for correcting this distortion by immediately following each image scan with a rescan of a small, narrow portion of the same area with the slow and fast scan axes reversed. The original, full image is corrected using a low-order polynomial mapping function, with coefficients determined by a pixel-wise comparison between the original full and rescanned partial images. We demonstrate here that this method can correctly remove distortion from a wide range of images with a precision of better than one pixel, and is also robust to common imaging artifacts. We also address some of the programming considerations that have gone into implementing this computationally intensive technique, which can now be performed using standard desktop hardware in times that range between a few seconds and a few minutes.
扫描探针显微镜(SPM)图像中的常见失真源是“热漂移”,即在扫描过程中温度的微小变化会导致样品和显微镜中不同材料的缓慢热膨胀。我们在这里描述了一种通过在每次图像扫描后立即用慢扫描和快扫描轴反转的相同区域的小窄部分的重新扫描来纠正这种失真的方法。使用低阶多项式映射函数对原始全图像进行校正,其系数由原始全图像和重新扫描的部分图像之间的逐像素比较确定。我们在这里证明,该方法可以以优于一个像素的精度正确地去除各种图像的失真,并且对常见的成像伪影也具有鲁棒性。我们还讨论了实现这种计算密集型技术所涉及的一些编程注意事项,现在可以在几秒钟到几分钟的时间内使用标准台式硬件来执行该技术。