摘要
针对传统电子显微镜采集细胞图像背景模糊、导致显微镜自动聚焦效率和数据采集实时性降低的问题,设计一个在无线环境下的微型荧光显微镜采集系统。通过该系统获取图像原始信息后,提出基于DeblurGAN的离焦图像去模糊方法,在DeblurGAN网络中加入ResNet残差块和入BAM注意力机制,对显微镜下的细胞边缘进行修复处理,从而削弱背景,并增强图像清晰度;通过该系统实现电子显微镜数据实时采集。实验结果表明,DeBlurGAN+BAM方法进行图像去模糊所用时间仅为0.227 ms,远远低于其他去模糊方法,本方法的PSNR和SSIM两个指标均保持在第一位,由此说明采用本方法能够增强图像清晰度、减少显微镜的聚焦时间,从而提高显微镜聚焦效率,进一步缩短数据采集时间,提高数据采集实时性。
A microfluorescence microscope acquisition system is designed for the problem of fuzzy background of traditional electron microscope cell image acquisition and reduced real-time data acquisition in wireless environment.After obtaining the original image information through the system, the DeblurGAN-based image deblur method is proposed, adding ResNet residual block and BAM attention mechanism in the DeblurGAN network to repair the cell edge under the microscope, weaken the background and enhance image clarity, and real-time electron microscope data collection is realized through the system.The experimental results show that the DeBlurGAN + BAM method takes only 0.227 ms, which is far lower than other methods. This method remains in the first place in PSNR and SSIM, which shows that this method can enhance the image clarity, reduce the focus time of the microscope, thus improve the microscope focus efficiency, further shorten the data acquisition time and improve the real-time of data acquisition.
作者
窦敏娜
王晓霞
DOU Minna;WANG Xiaoxia(Xianyang Vocational Technical College,Xianyang Shanxi 712000,China)
出处
《自动化与仪器仪表》
2023年第2期51-55,共5页
Automation & Instrumentation
基金
2021年全国轻工职业教育教学指导委员会课题《高职生物化学课程思政教学改革研究与实践》(QGHZW2021134)
陕西省教育科学“十四五”规划2021年度课题《高职生物化学课程混合式教学的研究与实践》(SGH21Y0593)。
