基于概率神经网络的寄生虫卵显微图像识别

Recognition of Images of Parasite Egg Based on Probabilistic Neural Network

病原体(虫卵)检测是诊断寄生虫病的最常用和最可靠的方法。该文对寄生虫卵显微图像的自动识别进行了研究,设计了一个基于概率神经网络的分类器。通过对血吸虫等9种寄生虫卵的显微图像进行自动识别,取得了平均正确识别率为99.23%的较好结果。

The detection of pathogeny(parasite eggs) is the most common and reliable way of diagnosis of parasite in-fection.In this article,a research of automatic recognition of images of parasite egg is performed.With the classifier based on probabilistic neural network,through the automatic recognition of nine kinds of images of parasite egg,a satisfactory recognition rate of 99.23% is obtained.