群智能算法的激光共焦扫描显微图像分类研究

Study on laser confocal scanning microscopic image classification based on swarm intelligence algorithm

为了提升图像分类效果,研究群智能算法的激光共焦扫描显微图像分类方法。依据激光共焦扫描显微镜成像原理,获取显微图像,并分成训练图像集与预测图像集。利用蚁群算法,提取两个图像集的特征向量,通过训练图像集的特征向量,训练交叉核支持向量机,建立支持向量机模型。依据时变函数调整蚁群算法内信息素的更新规则,改进蚁群算法。利用改进蚁群算法,优化支持向量机模型的参数,获取最优激光共焦扫描显微图像分类模型,在分类模型内输入预测显微图像集的特征向量,输出激光共焦扫描显微图像分类结果。实验证明:该方法可有效分类激光共焦扫描显微图像,具备较高的图像分类精度。在不同放大倍数下,图像误分类率较低,分类后图像的平均梯度值较高,即图像清晰度高。

In order to improve the effect of image classification,a laser confocal scanning microscopic image classification method based on swarm intelligence algorithm is studied.According to the imaging principle of laser confocal scanning microscopes,the microscopic image is obtained and divided into a training image set and a prediction image set.Using an ant colony algorithm,the feature vectors of two image sets are extracted.By training the feature vectors of image sets,the cross kernel support vector machine is trained,and the support vector machine model is established.According to the time-varying function,the pheromone update rules in the ant colony algorithm are adjusted to improve the ant colony algorithm.The improved ant colony algorithm is used to optimize the parameters of the support vector machine model to obtain the optimal laser confocal scanning micro image classification model.The feature vector of the predicted micro image set is input into the classification model,and the laser confocal scanning micro image classification results are output.Experiments show that this method can effectively classify laser confocal scanning micro images and has high image classification accuracy.Under different magnification,the image misclassification rate is low,and the average gradient value of the classified image is high,that is,the image definition is high.