扫描相干衍射成像中位置误差校正研究进展OA
Research progress in position error correction for scanning coherent diffractive imaging
作为相干衍射成像技术实现方式中的一种,扫描相干衍射成像在一定程度上克服了传统的相干衍射成像算法收敛速度慢、重建停滞、重建结果不唯一等问题.但是其本身也存在着一些问题,如位置误差、噪声干扰、光源非完全相干、成像耗时长等.其中,位置误差问题是扫描相干衍射成像相对于其他显微成像技术所特有的,显著影响了其成像质量.本文针对这一影响因素,首先介绍了扫描相干衍射成像背景信息以及影响重建质量的各种因素,然后介绍了成像过程中的位置误差校正方法和深度学习在这一方向的应用,以及对于误差校正效果的各种评判方法,最后总结并展望了其未来发展方向.这为扫描相干衍射成像质量提升方法的研究提供了参考,并对其在X射线成像领域的应用具有重要的指导意义.
As an implementation approach for coherent diffraction imaging(CDI)technology,scanning coherent diffraction imaging(SCDI)has partially overcome the limitations of traditional CDI algorithms,such as slow convergence rate,reconstruction stagnation,and non-unique reconstruction solutions.However,it still suffers from inherent drawbacks and challenges,including position errors,noise interference,incomplete source coherence,and long imaging duration.Among these issues,position errors,which are unique to SCDI compared with other microscopic imaging techniques,significantly degrade the final imaging quality.Focusing on this critical factor,this paper first presents the research background of SCDI and summarizes various factors affecting reconstruction quality.Subsequently,it elaborates on position error correction methods in the imaging workflow,the application of deep learning in this field,and various evaluation methods for error correction performance.Finally,this paper summarizes the current progress and prospects the future development trends.This work provides a reference for exploring imaging quality improvement methods of SCDI,and is of important guiding significance for its applications in X-ray imaging.
张文采;许子健
中国科学院上海应用物理研究所 上海 201800||中国科学院大学 北京 100049||中国科学院上海高等研究院 上海同步辐射光源 上海 201204中国科学院上海应用物理研究所 上海 201800||中国科学院大学 北京 100049||中国科学院上海高等研究院 上海同步辐射光源 上海 201204
能源科技
扫描相干衍射成像局部位置误差全局位置误差深度学习
Scanning coherent diffraction imagingLocal position errorGlobal position errorDeep learning
《辐射研究与辐射工艺学报》 2026 (4)
14-34,21
国家重点研发计划(2022YFA1603702,2021YFA1601001)和国家自然科学基金(12335019,11875316) Supported by National Key Research and Development Program of China(2022YFA1603702,2021YFA1601001)and the National Natural Science Foundation of China(12335019,11875316)
评论