Synchrotron Radiation Workshop (SRW) is a powerful synchrotron radiation simulation tool and has been widely used at synchrotron facilities all over the world. During the last decade, many types of X-ray wavefront sensors have been developed and used. In this work, we present our recent effort on the development of at-wavelength metrology simulation based on SRW mainly focused on the Hartmann Wavefront Sensor (HWS). Various conditions have been studied to verify that the simulated HWS is performing as expected in terms of accuracy. This at-wavelength metrology simulation tool is then used to align KB mirrors by minimizing the wavefront aberrations. We will present our optimization process to perform an ‘in situ’ alignment using conditions as close as possible to the real experiments (KB mirrors with different levels of figure errors or different misalignment geometry).
Phase Measuring Deflectometry (PMD) is a powerful tool to measure the three-dimensional shape for freeform specular surfaces. In this work, a model based method is applied to PMD, called as Modal Phase Measuring Deflectometry (MPMD). The surface height and slopes are represented in mathematical models and updated by optimizing the model coefficients, in order to minimize the discrepancy between the reprojection in ray tracing and the actual measurement. The pose of the screen relative to the camera is pre-calibrated and then optimized together with the surface shape coefficients. Moreover, the correspondence residuals because of the discrepancies between the modal estimation and practical acquisition are analyzed. Slope residuals are calculated from these discrepancies. Zonal integration methods which are good at dealing with local variations are used to reconstruct the height residual for compensation. Simulations and experiments are conducted to demonstrate the feasibility of the proposed approach.
After years of development from a concept to early experimental stage, X-ray Deformable Mirrors (XDMs) are used in many synchrotron/free-electron laser facilities as a standard x-ray optics tool. XDM is becoming an integral part of the present and future large x-ray and EUV projects and will be essential in exploiting the full potential of the new sources currently under construction. The main objective of using XDMs is to correct wavefront errors or to enable variable focus beam sizes at the sample. Due to the coupling among the <i>N</i> actuators of a DM, it is usually necessary to perform a calibration or training process to drive the DM into the target shape. Commonly, in order to optimize the actuators settings to minimize slope/height errors, an initial measurement need to be collected, with all actuators set to 0, and then either <i>N</i> or <i>2N</i> measurements are necessary learn each actuator behavior sequentially. In total, it means that <i>N</i>+1 or <i>2N</i>+1 scans are required to perform this learning process. When the actuators number <i>N</i> is important and the actuator response or the necessary metrology is slow then this learning process can be time consuming. In this work, we present a fast and accurate method to drive an x-ray active bimorph mirror to a target shape with only 3 or 4 measurements. Instead of sequentially measuring and calculating the influence functions of all actuators and then predicting the voltages needed for any desired shape, the metrology data are directly used to “guide” the mirror from its current status towards the particular target slope/height via iterative compensations. The feedback for the iteration process is the discrepancy in curvature calculated by using B-spline fitting of the measured height/slope data. In this paper, the feasibility of this simple and effective approach is demonstrated with experiments.
Nowadays, 3D face recognition has become a subject of considerable interest in the security field due to its unique advantages in domestic and international. However, acquiring color-textured 3D faces data in a fast and accurate manner is still highly challenging. In this paper, a new approach based on color speckle projection for 3D face data dynamic acquisition is proposed. Firstly, the projector-camera color crosstalk matrix that indicates how much each projector channel influences each camera channel is measured. Secondly, the reference-speckle-sets images are acquired with CCD, and then three gray sets are separated from the color sets using the crosstalk matrix and are saved. Finally, the color speckle image which is modulated by face is captured, and it is split three gray channels. We measure the 3D face using multi-sets of speckle correlation methods with color speckle image in high-speed similar as one-shot, which greatly improves the measurement accuracy and stability. The suggested approach has been implemented and the results are supported by experiments.
The study of 3D shape measurement by digital speckle temporal sequence correlation have drawn a lot of attention by its own advantages, however, the measurement mainly for depth z-coordinate, horizontal physical coordinate (x, y) are usually marked as image pixel coordinate. In this paper, a new approach for the system calibration is proposed. With an auxiliary camera, we made up the temporary binocular vision system, which are used for the calibration of horizontal coordinates (mm) while the temporal sequence reference-speckle-sets are calibrated. First, the binocular vision system has been calibrated using the traditional method. Then, the digital speckles are projected on the reference plane, which is moved by equal distance in the direction of depth, temporal sequence speckle images are acquired with camera as reference sets. When the reference plane is in the first position and final position, crossed fringe pattern are projected to the plane respectively. The control points of pixel coordinates are extracted by Fourier analysis from the images, and the physical coordinates are calculated by the binocular vision. The physical coordinates corresponding to each pixel of the images are calculated by interpolation algorithm. Finally, the x and y corresponding to arbitrary depth value z are obtained by the geometric formula. Experiments prove that our method can fast and flexibly measure the 3D shape of an object as point cloud.