摘要
为了保证眼底测量结果的准确性、客观性、可重复性以及实用性,提出了彩色眼底图像自动分割与定量分析的算法。具体步骤如下:首先对彩色视网膜血管图像进行网格划分,其次对包含重要血管信息的网格区域实现Otsu阈值分割,在此基础上对其它相邻网格进行区域生长算法分割,最后由计算机统一处理得到视网膜血管的网络径线。实验结果表明:该算法提取的血管网络径线连续性较好,血管中心线定位准确,抗干扰能力较强,处理速度较快,具有较高的临床应用价值。
segmentation algorithm of color retinal vascular images is proposed in this article to guarantee the accuracy, objectivity, reproducibility and practicability of retinal image measurement. The procedures were manipulated as following： firstly the green color image was extracted from RGB image and plotted onto grids, secondly the grid regions that contain important vascular information were segmented by using Otsu threshold, then regiongrowing algorithm was applied in neighbor grid regions, lastly retinal vascular image framework was extracted to achieve image parameter measurement. Experimental results demonstrate that the performance of this algorithm is superior to that of traditional methods and this algorithm can achieve fast retinal image measurement with high accuracy on a common configuration personal computer.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2007年第7期1281-1285,共5页
Chinese Journal of Scientific Instrument
关键词
视网膜血管图像
人体生物特征
图像分割
计算机图像处理
retinal vascular image
human biologic characteristic
image segmentation
computer-aided image processing
