角膜老年环检测方法研究

当人体脂类物质代谢发生异常时,会在角膜处形成角膜老年环。角膜老年环主要呈云雾环状覆盖在虹膜外环中,角膜老年环的检测可以作为一种辅助手段,对及时方便了解体内脂类代谢异常状况具有重要意义。针对目前方法中由于没有对角膜老年环边界精确检测而造成的鲁棒性不高,以及当存在K-F环时会发生检测错误的问题,提出了一种角膜老年环检测方法。首先,对其彩色图像进行分割与归一化;其次,利用基于梯度响应最优算法提取环状特征内外边界,通过定义的光照影响检测模型提高算法对光照的鲁棒性;最后,通过定义的宽度特征量化算子以及颜色特征量化算子完成角膜老年环的检测。实验结果表明,在采集的1 968幅图像中该方法能够达到96.3%的识别率,算法不仅不受光照影响而且当患者具有K-F环时,本文方法仍能够取得很好的检测效果。

Research on corneal arcus detection method

When the abnormality of lipid metabolism in human body happens, blood lipid will deposit in the cornea, this kind of symptom is called corneal arcus. The corneal arcus mainly covers on the outer ring of the iris and appears as cloudy ring shape; the detection of the corneal arcus can be used as an auxiliary method for diagnosis, which has an important significance for revealing the abnormal lipid metabolism in human body timely. Traditional methods can not accurately detect the boundary of the corneal arcus, which causes the problem of poor robustness, and erroneous detection may occurs when the iris image has K F ring. In this paper, a corneal arcus detection method is proposed. Firstly, the segmentation and normalization of the iris color image are conducted. Secondly, the optimal algorithm based on gradient response is used to extract the inner and outer boundaries of the ring characteristics of the corneal arcus; with the defined illumination influence detection model, the robustness of the algorithm to illumination is improved. Finally, the defined width feature and color feature quantization operators are used to achieve the corneal arcus detection of the iris image. 1968 iris images in our database were analyzed; and experiment results show that with the proposed method the classification accuracy reaches 96.3%. The proposed method is not sensitive to illumination and still can achieve good detection performance when the patient has K F ring.