Crystalline lens turbidity degree detection method based on convolutional neural network

The invention discloses a crystalline lens turbidity detection method based on a convolutional neural network. The method comprises the following steps: (1) preprocessing a to-be-detected image by adopting an illumination enhancement method to form preprocessed image data; and (2) substituting the preprocessed image data in the step (1) into a crystalline lens turbidity detection learning model torealize detection. According to the invention, the Inception-V3 model and parameters which are pre-trained by the ImageNet are utilized. The training is carried out by adopting the thought of transfer learning to obtain the classification model. After the system is completed, crystalline lens turbidity research and comparison can be carried out in real time through a mobile phone APP. 基于卷积神经网络的晶状体浑浊程度检测方法，该方法步骤如下：（1）、对待测图像采用光照增强法进行预处理形成预处理图像数据；（2）、将（1）步骤中的预处理图像数据代入晶状体浑浊程度检测学习模型实现检测，本发明利用ImageNet预训练过的Inception-V3模型及参数，并采用迁移学习的思想进行训练从而得到分类模型，该系统完成后可实现通过手机APP实时进行晶状体浑浊研究对比。