一种快速精确的瞳孔和角膜反射光斑中心定位算法的研究

发布时间：2018-06-15 06:30

本文选题：定位算法 + 瞳孔中心　；参考：《中国生物医学工程学报》2017年04期

【摘要】：眼动追踪系统在临床医学中有着重要应用,为提高系统中眼图特征值的识别率,创新性地提出一种双层过滤的瞳孔中心和角膜反射光斑中心的定位算法。瞳孔中心定位算法首先对获取的眼睛图像二值化处理,然后以100~300为阈值设定轮廓点数,进行第一次瞳孔过滤,最后用最小二乘法椭圆拟合瞳孔轮廓,并将最接近圆的轮廓作为瞳孔轮廓,继而得到瞳孔中心。在此基础上,进一步对瞳孔中心一定范围内的眼图进行边缘检测,求其轮廓,并找出轮廓的外接矩形中心,然后以瞳孔中心为基点进行坐标转换,最终求出角膜反射光斑中心。算法在一定程度上缩小图像处理的区域,提高实时性和反射光斑的识别率。实验结果表明,所设计的算法能够实时地对不同情况下的瞳孔和角膜反射光斑中心进行准确定位,瞳孔-角膜反射向量坐标最大均方差低于0.74像素,平均误差低于0.53像素。算法运行速度达到12.8 ms/帧。算法满足精确性的同时具有很强的鲁棒性,为后续眼动设备的开发提供重要的参考价值。
[Abstract]:Eye movement tracking system plays an important role in clinical medicine. In order to improve the recognition rate of eye image eigenvalues, a double-layer filtering algorithm for locating the center of pupil and corneal reflection spot is proposed. The pupillary center location algorithm first binarizes the acquired eye image, then sets the contour points with a threshold of 100 ~ 300, carries on the first pupil filtering, and finally uses the least square ellipse to fit the pupil contour. The closest contour to the circle is used as the pupil contour, and then the pupil center is obtained. On this basis, the edge of the eye image in a certain range of pupil center is further detected, the contour is obtained, and the external rectangular center of the contour is found, then the center of the corneal reflective spot is obtained by coordinate transformation based on the center of the pupil. The algorithm reduces the area of image processing to a certain extent and improves the real time and the recognition rate of reflected spot. Experimental results show that the proposed algorithm can accurately locate the center of pupil and corneal reflection spot in real time. The maximum mean square deviation of pupillary and corneal reflection vector coordinates is less than 0.74 pixels and the average error is less than 0.53 pixels. The speed of the algorithm is 12.8 ms/ frame. The algorithm not only satisfies the accuracy but also has strong robustness, which provides an important reference value for the subsequent development of eye movement equipment.
【作者单位】：重庆大学生物医学工程学院;重庆市医疗电子工程技术研究中心;中国生物医学工程学会;
【基金】：国家科技支撑计划项目(2013BAI03B04&2015BAI01B14)
【分类号】：R770.4;TP391.41

【相似文献】