3-D <i>μ</i>XRF Imaging System for Curved Surface

摘要

Micro X-ray fluorescence (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> XRF) is a technology for nondestructive detection of elements with high resolution. Currently, the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> XRF imaging system cannot scan with a fixed distance and incident angle, which highly affects the measurement consistency of curved objects. To solve this problem, we developed a novel 3-D <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> XRF imaging system based on a robotic arm and a depth camera. In our system, the depth camera is used to capture both the coordinates and normal vectors of the contour of the object, while the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> XRF spectrometer is installed on the six-axis robotic arm to flexibly scan the curved surface according to the contour. Compared to the existing systems, the proposed 3-D <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> XRF imaging system eliminates the interference to the counts caused by the changes in the X-ray incident angle and distance, which improves the 3-D imaging quality of curved surfaces. The highest resolution of the system can be controlled to within <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$50~\mu $ </tex-math></inline-formula>m, and the single-point acquisition time can be as short as milliseconds. The alloy sample scanning experiment quantitatively demonstrated the fluctuation of the detector counts is only 0.62% within 18 mm of the alloy sample surface, while the fluctuations of the other two scanning methods are 3.79% and 9.42%. The oil painting experiment validated the accuracy of the imaging system in element identification and imaging, by comparing it with ground truth. Teapot and cylinder experiments are performed to demonstrate the system’s capability for high-precision and high-resolution 3-D imaging of objects in any orientation.