应用于眼底成像的可见光瞳孔定位与对准方法

邢利娜; 孔文; 唐宁; 陈一巍; 史国华; 何益

doi:10.37188/CO.2024-0065

应用于眼底成像的可见光瞳孔定位与对准方法

doi: 10.37188/CO.2024-0065

cstr: 32171.14.CO.2024-0065

邢利娜^{1, 2,},
孔文^1,,
唐宁^1,,
陈一巍^2,,
史国华^2, ,,
何益^2, ,

1.
中国科学技术大学生物医学工程学院(苏州), 江苏苏州 215163
2.
中国科学院苏州生物医学工程技术研究所江苏省医用光学重点实验室, 江苏苏州 215163

基金项目: 国家重点研发计划（No. 2021YFC2401401）；国家自然科学基金项目（No. 62075235）；中国科学院青年创新促进会（No. 2019320）；中国科学院战略性先导科技专项（No. XDA16021304）

详细信息

作者简介:
邢利娜（1985—），女，河南洛阳人，博士研究生，副研究员，2010年于重庆大学获得硕士学位，主要从事光学成像、高分辨率眼科成像与检测等方面的研究。E-mail：xingln@sibet.ac.cn

史国华（1981—），男，浙江湖州人，博士，研究员，博士生导师，2011年于中国科学院光电技术研究所获得博士学位，主要从事高分辨率眼科成像与检测、在体光学超分辨成像、活体视觉神经网络等方面的研究。E-mail：shigh@sibet.ac.cn

何　益（1984—），男，四川营山人，博士，研究员，博士生导师，2013年于中国科学院光电技术研究所获得博士学位，主要从事波前处理、眼科光学、生物光子学、基于智能计算的精准医疗等方面的研究。E-mail：heyi@sibet.ac.cn

中图分类号: TP23;TH786
计量
- 文章访问数: 159
- HTML全文浏览量: 86
- PDF下载量: 36
- 被引次数: 0
出版历程
- 收稿日期: 2024-04-07
- 修回日期: 2024-04-30
- 录用日期: 2024-05-09
- 网络出版日期: 2024-05-22

Visible light pupil localization and alignment method for fundus imaging

XING Li-na^{1, 2
,},
KONG Wen^1
,,
TANG Ning^1
,,
CHEN Yi-wei^2
,,
SHI Guo-hua^{2
, ,},
HE Yi^{2
, ,}

1.
School of Biomedical Engineering (Suzhou), University of Science and Technology of China, Suzhou, Suzhou, 215163, China
2.
Jiangsu Key Laboratory of Medical Optics, Suzhou Institute of Biomedical Engineering and Technology, Chinese Academy of Science, Suzhou, 215163, China

Funds: Supported by National Key Research and Development Program of China (No.2021YFC2401401); National Natural Science Foundation of China (No. 62075235); Youth Innovation Promotion Association, CAS (No.2019320); Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDA16021304)

More Information

Corresponding author: shigh@sibet.ac.cn; heyi@sibet.ac.cn

摘要

摘要:
为了降低眼底成像过程对操作人员水平的依赖，提出了一种采用可见光瞳孔成像的人眼瞳孔自动快速定位与对准方法。采用可见光摄像模组和三维电动位移台，在实验室眼底成像系统上搭建瞳孔对准装置。用霍夫梯度法提取图像的有效区域，得到眼底成像系统的中心。利用最大类间方差法和图像直方图特征得到了瞳孔区域后，利用最小圆拟合方法确定瞳孔中心。通过反馈控制电动位移台的移动，实现眼底成像系统中心与瞳孔中心对准。实验结果表明：人眼瞳孔的平均识别速度为0.11 s，瞳孔中心的识别平均准确率达98.7%，中心偏差的平均欧氏距离为4.3个像素。上述结果满足眼底成像系统的实时性和准确性要求，为眼底成像系统提供了一种高效的自动瞳孔对准解决方案。
- 眼底成像 /
- 瞳孔定位 /
- 自动对准 /
- 可见光
Abstract:
To mitigate reliance on operators during fundus imaging, an automated rapid localization and alignment method for the human pupil using visible light pupil imaging was proposed. Initially, the pupil alignment device was constructed on a laboratory fundus imaging system using a visible light camera module and a three-dimensional electric displacement stage. Subsequently, the effective area of the image was extracted using the Hough gradient method to determine the center of the fundus imaging system. The pupil region was identified through the maximum inter-class variance method and image histogram feature, while the center of the pupil was ascertained via the minimum circle fitting method. Ultimately, the electric displacement stage's movement is regulated through feedback mechanisms, ensuring that the center of the fundus imaging system aligns precisely with the pupil's center. The experimental results show that the average recognition speed of human pupil is 0.11 s, the average recognition accuracy of the pupil center is 98.7%, and the average Euclidean distance of the center deviation is 4.3 pixels. It can satisfy the system requirements of the real-time and accuracy, and provides an efficient automatic pupil alignment solution for fundus imaging system.
- fundus imaging /
- pupil localization /
- automatic alignment /
- visible light

HTML全文

图 1 瞳孔对准装置

Figure 1. The pupil alignment device

下载: 全尺寸图片幻灯片

图 2 瞳孔对准流程图

Figure 2. Flow chart of pupil alignment

下载: 全尺寸图片幻灯片

图 3 Sobel算子矩阵

Figure 3. The Sobel operator matrix

下载: 全尺寸图片幻灯片

图 4 成像系统中心定位的流程图

Figure 4. Flow chart of imaging system center localization

下载: 全尺寸图片幻灯片

图 5 随机选取的4张图片的成像系统中心、原始图像及有效区域

Figure 5. The center of the imaging system, the original image, and the effective region for four randomly selected images

下载: 全尺寸图片幻灯片

图 6 4组图像的有效区域和核心区域图像的直方图

Figure 6. The histograms of the effective area and the core region for the four images

下载: 全尺寸图片幻灯片

图 7 瞳孔中心定位的流程图

Figure 7. Flow chart of pupil center localization

下载: 全尺寸图片幻灯片

图 8 （a）形态学运算和（b）瞳孔中心的图像

Figure 8. Images of (a) morphological operations and (b) pupil center

下载: 全尺寸图片幻灯片

图 9 采用最大类间方差法的图像。（a）成像系统中心；（b）核心区域；（c）二值化；（d）形态学；（e）瞳孔中心

Figure 9. Images obtained by maximum inter-class variance method. (a) The center of imaging system; (b) core region; (c) binarization results; (d) morphological results; (e) pupil center

下载: 全尺寸图片幻灯片

图 10 采用直方图谷值识别瞳孔中心。（a）成像系统中心；（b）核心区域；（c）再次采用最大类间方差法得到结果；（d）二值化；（e）形态学；（f）瞳孔中心

Figure 10. Pupil center determined by histogram valley method. (a) The center of the imaging system; (b) core region; (c) using twice maximum between-cluster results; (d) binarization results; (e) morphological results; (f) pupil center

下载: 全尺寸图片幻灯片

图 11 成像系统中心和瞳孔中心的定位结果

Figure 11. Localization results of imaging system center and pupil center

下载: 全尺寸图片幻灯片

图 12 两种方法与手工标记的瞳孔中心像素偏差的欧氏距离

Figure 12. Euclidean distance of deviation of pupil center located by the proposed algorithm and circle approximation algorithm from manually marked pupil center

下载: 全尺寸图片幻灯片

图 13 原始图像及瞳孔对准后的图像结果。（a）原始图像及（b）其瞳孔定位结果；（c）对准后的原始图像及其（d）瞳孔定位结果

Figure 13. The original image and the pupil-aligned images. (a) The original image and (b) it’s pupil localization results; (c) the aligned original image and (d) it’s pupil localization results

下载: 全尺寸图片幻灯片

参考文献(21)

[1]	汪路涵, 巩岩, 张艳微, 等. 人眼自动对焦及瞳孔中心自动对准系统[J]. 光学学报,2023,43(5):0511002. doi: 10.3788/AOS221703 WANG L H, GONG Y, ZHANG Y W, et al. Human eye autofocus and pupil center auto-alignment system[J]. Acta Optica Sinica, 2023, 43(5): 0511002. (in Chinese). doi: 10.3788/AOS221703
[2]	王鹏, 温宏韬, 王世龙. 基于近眼红外图像的高精度瞳孔中心定位方法[J]. 哈尔滨理工大学学报,2022,27(5):38-46. WANG P, WEN H T, WANG SH L. High-precision pupil center positioning method based on near-eye infrared image[J]. Journal of Harbin University of Science and Technology, 2022, 27(5): 38-46. (in Chinese).
[3]	蔡怀宇, 史玉, 娄世良, 等. 应用于红外眼科疾病检测的瞳孔定位算法[J]. 中国光学,2021,14(3):605-614. doi: 10.37188/CO.2020-0170 CAI H Y, SHI Y, LOU SH L, et al. Pupil location algorithm applied to infrared ophthalmic disease detection[J]. Chinese Optics, 2021, 14(3): 605-614. (in Chinese). doi: 10.37188/CO.2020-0170
[4]	王晶, 高峰, 李婉越, 等. 瞳孔中心点自动定位与对准装置[J]. 光学精密工程,2019,27(6):1370-1377. doi: 10.3788/OPE.20192706.1370 WANG J, GAO F, LI W Y, et al. Device for automatic pupil center location and alignment[J]. Optics and Precision Engineering, 2019, 27(6): 1370-1377. (in Chinese). doi: 10.3788/OPE.20192706.1370
[5]	CHEN G ZH, DONG ZH H, WANG J, et al. Pupil localization algorithm based on improved U-Net network[J]. Electronics, 2023, 12(12): 2591. doi: 10.3390/electronics12122591
[6]	XUE P X, WANG CH Y, HUANG W B, et al. Pupil Centre's localization with transformer without real pupil[J]. Multimedia Tools and Applications, 2023, 82(16): 25467-25484. doi: 10.1007/s11042-023-14403-3
[7]	王鉴, 张荣福. 基于可变形卷积的单帧图像眼球定位追踪[J]. 光学仪器,2021,43(6):26-31. WANG J, ZHANG F R. Single-frame image eyeball tracking based on deformable convolution[J]. Optical Instruments, 2021, 43(6): 26-31. (in Chinese).
[8]	张贺童, 姚康, 裴融浩, 等. 基于ADEU-Net分割网络的瞳孔精确分割方法[J]. 计算机工程与应用,2023,59(2):212-221. ZHANG H T, YAO K, PEI R H, et al. Accurate pupil segmentation based on ADEU-Net segmentation network[J]. Computer Engineering and Applications, 2023, 59(2): 212-221. (in Chinese).
[9]	张宏薇, 王仕洋, 李宪龙, 等. 基于Hough变换的瞳孔识别方法研究与实现[J]. 液晶与显示,2016,31(6):621-625. doi: 10.3788/YJYXS20163106.0621 ZHANG H W, WANG SH Y, LI X L, et al. Research and implementation of pupil recognition based on Hough transform[J]. Chinese Journal of Liquid Crystals and Displays, 2016, 31(6): 621-625. (in Chinese). doi: 10.3788/YJYXS20163106.0621
[10]	ABDULLAH M A M, DLAY S S, WOO W L. Fast and accurate pupil isolation based on morphology and active contour[J]. International Journal of Information and Electronics Engineering, 2014, 4(6): 418-422.
[11]	JAN F, USMAN I, AGHA S. Iris localization in frontal eye images for less constrained iris recognition systems[J]. Digital Signal Processing, 2012, 22(6): 971-986. doi: 10.1016/j.dsp.2012.06.001
[12]	陈静瑜, 林丽媛, 刘冠军, 等. 一种基于几何形状特征的实时瞳孔定位追踪技术[J]. 天津科技大学学报,2021,36(3):65-71. CHEN J Y, LIN L Y, LIU G J, et al. A real-time pupil positioning and tracking technology based on geometric shape features[J]. Journal of Tianjin University of Science & Technology, 2021, 36(3): 65-71. (in Chinese).
[13]	WAN ZH H, XIONG C H, CHEN W B, et al. Robust and accurate pupil detection for head-mounted eye tracking[J]. Computers & Electrical Engineering, 2021, 93: 107193.
[14]	童言, 穆平安, 林玉莹. 基于圆形滑动窗口的瞳孔定位算法[J]. 电子测量技术,2021,44(2):99-103. TONG Y, MU P A, LIN Y Y. Pupil location algorithm based on circular sliding window[J]. Electronic Measurement Technology, 2021, 44(2): 99-103. (in Chinese).
[15]	MIN-ALLAH N, JAN F, ALRASHED S. Pupil detection schemes in human eye: a review[J]. Multimedia Systems, 2021, 27(4): 753-777. doi: 10.1007/s00530-021-00806-5
[16]	JAN F, USMAN I, AGHA S. Reliable iris localization using Hough transform, histogram-bisection, and eccentricity[J]. Signal Processing, 2013, 93(1): 230-241. doi: 10.1016/j.sigpro.2012.07.033
[17]	KHAN T M, KHAN M A, MALIK S A, et al. Automatic localization of pupil using eccentricity and iris using gradient based method[J]. Optics and Lasers in Engineering, 2011, 49(2): 177-187. doi: 10.1016/j.optlaseng.2010.08.020
[18]	LIN ZH H, YU H F. The pupil location based on the OTSU method and Hough transform[J]. Procedia Environmental Sciences, 2011, 8: 352-356. doi: 10.1016/j.proenv.2011.10.055
[19]	ZHOU W J, LU X Y, WANG Y. A robust pupil localization via a novel parameter optimization strategy[J]. Wireless Communications and Mobile Computing, 2022, 2022: 2378911.
[20]	刘诗瑜, 赵夏冬, 温盼, 等. 基于FPGA的分段近似双边滤波算法设计与实现[J]. 液晶与显示,2024,39(10):1341-1349. Liu S Y, Zhao X D, Wen P. Design and implementation of piecewise approximation bilateral filtering algorithm based on FPGA[J]. Chinese Journal of Liquid Crystals and Displays, 2024, 39(10): 1341-1349.
[21]	黎明清, 王宇庆, 孙海江. 基于侧窗滤波改进的神经网络非均匀性校正算法[J]. 液晶与显示,2023,38(11):1567-1579. Li M Q, Wang Y Q, Sun H J. Improved neural network nonuniformity correction algorithm based on side window filter[J]. Chinese Journal of Liquid Crystals and Displays, 2023, 38(11): 1567-1579.